403(b), 457, and the TSP

Here is a fear that follows public servants into the staff lounge. Your friends in the corporate world got a 401(k) — a single plan, one menu, a match the company brags about in the offer letter, the whole thing wrapped up in one tidy enrollment screen. You got something that doesn't even look like the same animal. Instead of one plan you got a list of unfamiliar vendors — insurance companies, mostly, with names you've never heard — each selling something called a "tax-sheltered annuity," and a friendly salesperson who set up a folding table in the teachers' lounge at lunch, offering to "help you enroll" while you ate. And you couldn't tell, sitting there, whether you were being helped or hunted — whether the nice person across the table was on your side, whether the account they were steering you into was any good, or whether the whole arrangement was somehow second-class next to the clean 401(k) your corporate friends got handed.

So let's disarm that before we do anything else, because almost none of the fear survives contact with the facts. Your account is not second-class. A 403(b), a 457, and the federal Thrift Savings Plan are the 401(k)'s siblings — they run on the exact same engine. Same 2026 contribution limit ($24,500 of your own salary), same tax break, same decades of compounding that build a 401(k) into a retirement. If anything, the soldier or the federal worker reading this has it better than almost everyone in corporate America: the TSP is one of the lowest-cost retirement plans in the entire country, a plan a private-sector worker would envy. The thing that's true — and the only thing that's true — is that one of these accounts, the 403(b) sold by that table in the lounge, has a single real trap folded into it that a 401(k) usually doesn't. Name the trap and it stops being scary. It becomes a thing you sidestep.

That's what this lesson hands you: two skills, and only two. The first is reading the menu — telling the good, cheap option from the expensive one on a list that's often designed to make that hard. The second is sidestepping the one real trap, the high-fee annuity sold by a commissioned salesperson who, crucially, isn't required to put your interests first. Two skills, and the whole confusing list goes quiet. The salesperson at the table stops being a question mark. You walk up knowing exactly what you're looking for and exactly what to refuse.

This is the sibling of the two 401(k) lessons that came just before it — Part 1 (getting in, capturing the match, reading the menu) and Part 2 (living with the account over the decades). Everything you learned there about how an elective deferral works, how a match works, how an expense ratio quietly eats returns — all of it carries straight over, because these are the same machine in a different case. We won't re-teach the parts that are identical. We'll spend our time on what's genuinely different: the annuity trap to navigate, the superpowers some of these accounts have that a 401(k) doesn't, and the one plan that's flatly better than most 401(k)s in America.

Start with the reassuring truth that the rest of the lesson rests on: a 403(b), a governmental 457(b), and the Thrift Savings Plan are not exotic accounts. They are the 401(k) in different clothes. The defining move is identical to the one from the 401(k) lessons — your money leaves your paycheck and goes into investments before it ever reaches your bank account, and for a traditional contribution, before it's taxed. That contribution has the same name it had there: an elective deferral. Elective because you choose to make it; deferral because, with a traditional contribution, you're deferring the tax on that slice of income until you withdraw it decades from now. When Angela Morales — a 48-year-old public-school teacher in San Antonio, earning $58,000 a year — sends $200 a month into her school district's 403(b), that $200 is her elective deferral, working exactly as it would in a 401(k): it leaves the paycheck, lands in the retirement account, and (because hers is a traditional contribution) lowers her taxable income for the year. Same motion, set up once, mostly forgotten after.

The numbers are the same too, which is worth saying plainly because the fear in the intro was partly a fear of getting something smaller. You don't. For 2026, all three of these accounts share the very same elective-deferral limit as the 401(k): $24,500 of your own salary. (That figure is set by the IRS and nudges up most years with inflation, so it's worth re-checking each January — but for 2026 it's $24,500 across the board.) And the same age-50 catch-up applies: once you turn 50, you can put in an extra $8,000 on top, lifting your personal cap to $32,500. So Angela, at 48, is two years away from that extra $8,000 of room — and when she crosses 50, her 403(b) hands her the identical catch-up a 50-year-old in a corporate 401(k) gets. Not a smaller version. The same one. (We'll come back to a couple of catch-up wrinkles unique to these accounts later in the lesson; for now the headline is that the core limits match the 401(k) dollar for dollar.)

What differs isn't the engine — it's which door you walk through to reach it, and that's decided entirely by who employs you. Here is the map, the orientation that makes the rest of the lesson navigable:

A few things on that map deserve a sentence each, because they explain the lesson's whole shape. Notice the 403(b)'s "run by" line — the account is your employer's plan, but the actual money is held and managed by outside vendors, and that one structural fact is where the trouble lives. There's a clue in the account's other name: a 403(b) is also called a "tax-sheltered annuity," or TSA. That isn't a coincidence or a quirk. When Congress created the 403(b) in 1958, the law literally allowed only one thing inside it — annuities, which are insurance products. Plain mutual funds weren't even permitted until 1974. So for the first sixteen years of its existence, a 403(b) could hold nothing but insurance, and the insurance industry got there first and entrenched itself in the teachers'-lounge market. That history is why, decades later, the menu a teacher is handed is still so often dominated by insurance products — and why that friendly table at lunch is usually staffed by an insurance salesperson. We'll unpack exactly what that means, and how to step around it, in §2. For now, just hold the seed: the 403(b)'s annuity origins explain the one trap in this lesson.

One more framing, and it's the one that makes all three of these accounts make sense at once: every one of them is a supplement to a pension, not a replacement for it. This is the deepest difference from the corporate world. A typical 401(k) is the worker's whole retirement — there's no pension behind it. But Angela, as a Texas teacher, has the Teacher Retirement System pension (TRS) coming — a defined-benefit pension projected to pay her roughly $2,200 a month for life starting at age 62. Her 403(b) isn't meant to carry her retirement alone; it's meant to fill the gap the pension leaves. The same is true for a federal civilian, whose TSP sits on top of the FERS pension, and for a soldier like Tomás Rivera — a 38-year-old Army Sergeant First Class about two years from a 20-year retirement — whose TSP supplements the military pension he's earning. So when you size up one of these accounts, you're not asking "will this fund my whole old age?" You're asking "will this, plus my pension, get me where I need to be?" — a meaningfully gentler question, and one more reason these accounts are nothing to be afraid of.

That's the orientation. Same engine, three doors, each a supplement to a pension. The rest of the lesson is about the three places where the siblings stop being identical and each reveals its own signature feature — one to fear, two to celebrate. In the 403(b), it's the annuity trap: the high-fee insurance product sold by that lounge table, which §2 teaches you to navigate and escape. In the 457, it's a genuine superpower: a no-penalty escape hatch that lets you reach your money early without the 10% penalty every other retirement account charges — the early-retirement bridge Marisol Vega, a county public-health administrator planning to retire at 57, will use. And in the TSP, it's the gold standard: rock-bottom fees so low that fewer than one in a hundred funds in America can match them, which is exactly why a soldier's account can quietly outperform a corporate 401(k). One trap, two superpowers. We take them one at a time.

If the 401(k) is the private sector's retirement account, the 403(b) is the one waiting for you if you teach in a public school, work at a nonprofit hospital, or draw a paycheck from a church or a charity. Under the hood it is the same machine we just spent two lessons learning — you defer money from your paycheck before tax, it grows tax-deferred, the 2026 limits are identical — so almost everything from Lessons 16 and 17 carries straight over, and this lesson won't re-teach it. What it will teach is the one place the 403(b) is genuinely, dangerously different: the menu of things you're allowed to buy inside it.

That difference splits into two stories, and they need separating because they pull in opposite directions. One is a trap — the high-fee variable annuity that dominates many K-12 teachers' plans and quietly skims tens of thousands of dollars from a career of saving. The other is the reassurance that sits right next to it — that a good 403(b), the kind with a match and a low-cost index option, absolutely exists, is held by millions of people, and is every bit as good as a strong 401(k). We'll meet a teacher caught in the first story, and a teacher living the second, because telling them apart is the whole skill this section builds.

§2.1 — What a 403(b) is, and why it's full of annuities

A 403(b) is the teacher's and nonprofit worker's version of the 401(k): a retirement account your employer offers, into which you route money from your paycheck before it's taxed, where it grows tax-deferred until you withdraw it in retirement. The deferral mechanics are the same, the 2026 elective-deferral limit is the same $24,500, the age-50 catch-up is the same $8,000 — so if you understood the 401(k), you already understand 95% of the 403(b). Only certain employers can offer one: public schools and colleges, 501(c)(3) nonprofits, hospitals, and churches. A private for-profit company cannot — which is exactly why this is the account a public-school teacher meets instead of a 401(k).

Meet Angela Morales, who will anchor this whole section. Angela is 48, a public-school teacher in San Antonio, Texas, earning $58,000 a year (Texas has no state income tax, which will matter later). She has $34,000 in her 403(b) and adds $200 a month. She's roughly 14 years from the retirement age her Texas teachers' pension — the TRS pension — targets, age 62, at which point that pension is projected to pay her about $2,200 a month. The 403(b) is the supplement she's building on top of that pension, and how well it does over these next 14 years depends almost entirely on one decision she's about to face: what, exactly, her money is allowed to be invested in.

Here's where the 403(b) stops resembling the 401(k), and the reason is buried in a single tax-code letter. The IRS recognizes three different containers a 403(b) can be funded through, and they are not interchangeable. The first is the 403(b)(1) annuity contract — an insurance-company product, where your money buys into an annuity issued by an insurer. The second is the 403(b)(7) custodial account — money held at a bank or custodian and invested in mutual funds, the same kind of low-cost index funds Maya bought in her 401(k). The third is the 403(b)(9) retirement income account, available only to churches, which can hold either. Hold onto the first two, because the entire trap and the entire escape live in the difference between them: the 403(b)(1) annuity is the door that often costs you dearly, and the 403(b)(7) custodial account is the low-cost door most teachers are never told exists.

Why is a retirement account full of insurance annuities in the first place? History. When the 403(b) was created in 1958 it was annuity-only by law — which is why it's still sometimes called by its old name, the tax-sheltered annuity, or TSA. Mutual funds weren't even permitted inside it until 1974. So for sixteen years the only thing you could buy in a 403(b) was an insurance annuity, the insurance industry built its entire sales force around selling them to teachers, and that head start never went away. The entrenchment is still visible in the numbers: as of 2014, roughly 76% of all 403(b) assets were still sitting in annuities, not mutual funds. The account didn't fill up with annuities because annuities are best for teachers; it filled up because the insurers got there first and stayed.

There's a second structural fact that shapes Angela's situation, and it's easy to miss because it's an absence rather than a feature: her plan has no employer match. Many K-12 public-school 403(b) plans are purely voluntary salary-deferral programs — the district lets you contribute, but puts in nothing of its own. So unlike the 401(k), where the headline reason to participate was free match money, Angela has no match to capture. That changes the calculus: with no match offsetting costs, the fees on whatever she buys become the single most important number, because there's no employer contribution sitting on top to dwarf them. Every dollar of fee comes straight out of her own savings.

It would be unfair, though, to leave you thinking every 403(b) is matchless or predatory — so meet the counterexample now, because evenhandedness is the point. Marcus Williams, the Chicago high-school history teacher we'll build alongside through this lesson, has a genuinely good 403(b). He's 41, earns $68,000, and contributes 6% — $4,080 a year — and his district matches 3%, adding $2,040. That match is a guaranteed 50% return on his matched dollars, the same unbeatable deal the 401(k) lesson built itself around, and it exists inside a 403(b). Two teachers, two 403(b)s, two completely different deals — and the difference isn't the account type, it's what's offered inside it. A good 403(b) is a great account. The trap is specific, and Angela is closer to it than Marcus is.

One more piece of plumbing, because it's the form that actually puts money in motion. You don't enroll in a 403(b) the way you click through a 401(k) portal; you sign a salary reduction agreement, or SRA — the document you file with HR or the benefits office that tells payroll how much of your salary to divert into the plan and to which vendor. The SRA is the lever. It's also, as we'll see, the lever you pull to escape a bad annuity, because redirecting future contributions to a better vendor is done by signing a new SRA — which is free and takes effect on the next paycheck.

Finally, a protection difference worth stating plainly rather than burying as paperwork, because it's a real downside, not a technicality. Most governmental and church 403(b) plans — including Angela's public-school plan — are non-ERISA: they're exempt from the federal law (ERISA) that imposes fiduciary duties (a legal obligation to put your interests first) on private employer plans. In practice that means weaker federal oversight of the plan's costs and weaker federal creditor protection for the money inside it; recourse for a problem runs through state law rather than the federal Department of Labor. The flip happens the moment an employer contributes: a nonprofit plan with a match — like the hospital 403(b) held by Priya Williams — Marcus's wife, a Chicago RN earning $95,000 — whose hospital matches 4% — becomes ERISA-covered, gaining the fiduciary protections Angela's plan lacks. So Marcus's matched plan and Priya's matched plan carry a layer of protection Angela's voluntary, non-ERISA plan does not. It's not a reason to panic — state insurance regulation and the verification tools we'll cover still apply — but it is a genuine reason the menu inside a non-ERISA plan deserves the careful read the next section gives it.

§2.2 — The annuity trap: the fee stack that eats a teacher's retirement

The screen above is what Angela actually faces when she goes to enroll — and notice it isn't a single "pick your fund" menu like Maya's clean 401(k) screen. It's a vendor list: a dozen separate companies, each its own enrollment path, most of them insurance carriers selling variable annuities, with exactly one low-cost custodial index option tinted near the bottom. This is the K-12 403(b) experience that no other retirement account subjects you to, and learning to read it is the difference between Angela's two possible futures. Let's walk it the way she'd have to.

First, the term on most of those lines: a variable annuity. An annuity is an insurance contract; a variable annuity is one whose value rises and falls with investments — called sub-accounts — held inside the insurance wrapper. So a variable annuity inside a 403(b) is, in effect, mutual funds wearing an insurance overcoat, and you pay for the overcoat every year. (Annuities are a large topic with legitimate uses we treat fully in Lesson 30; here we need only the part that bites a teacher.) The people selling these on Angela's list are commissioned insurance agents — often the friendly rep who set up a table in the teachers' lounge during open enrollment — and the crucial thing about them is that they are not fiduciaries. They are not legally required to put Angela's interest ahead of their commission. That's not an accusation that any one of them is dishonest; it's a conflict of interest Angela has to manage herself, because no one in that room is obligated to manage it for her.

Now the part that actually costs money: the fees, which on a variable annuity don't arrive as one number but as a stack of them, layered so that no single layer looks alarming. Take a typical variable annuity on Angela's list, with an all-in cost of about 2.25% a year (illustrative, but squarely within real ranges we'll cite in a moment). That 2.25% is built from three charges piled on top of each other:

Those aren't invented numbers. In California, where a state law forced the costs into the open, 403(b) products were found to range from 0.56% all the way to 4.58% a year, averaging 1.78%; a separate Aon analysis put variable annuities specifically at an average of 2.25% — exactly the figure above. So 2.25% is not a worst case; it's a typical case. And the M&E charge at the top of the stack is the tell: it's the fee for the insurance wrapper, the thing that makes this an annuity rather than a plain mutual fund — and it's the cost that the low-cost custodial option simply doesn't have.

There's a second cost that doesn't show up as an annual percentage at all: the surrender charge. We met this term back in Lesson 5 — it's a penalty for taking your money out early — and annuities are where it lives most aggressively. A typical 403(b) annuity charges around 5% if you withdraw or move your money in the first several years, declining over a 5-to-7-year schedule until it reaches zero. The cruel detail is that each new contribution often starts its own fresh surrender clock — so even years in, Angela's most recent deposits can still be locked up. The money is never entirely free to leave, which is precisely what makes the wrapper sticky and the escape (next section) something to plan rather than just do.

Here's the argument that should settle it, and it's the one almost no salesperson volunteers. The single biggest selling point of an annuity is tax deferral — your money grows without being taxed each year. But a 403(b) is already tax-deferred. That's the entire point of the account. So putting a tax-deferred annuity inside a tax-deferred 403(b) buys you a shelter you already have, for free, a second time — and charges you the M&E fee for the privilege. It's like paying extra for waterproofing on a raincoat. The tax deferral isn't doing anything the account wasn't already doing; you're just paying the insurer's wrapper fee on top of it. That redundancy is the structural heart of why the high-fee variable annuity is a poor fit for this particular container, regardless of how the agent frames it.

Now watch what that fee stack actually does to Angela over the 14 years she has until age 62. Run her real numbers — $34,000 today plus $200 a month, at an illustrative 6% gross return every year — through both doors. In the ~2.25% variable annuity, her money nets about 3.75% a year and grows to $101,529. In the ~0.10% custodial index option, it nets about 5.90% and grows to $129,554. Same contributions, same market, same 14 years — the only difference is the fee — and the gap is $28,024:

Twenty-eight thousand dollars, and remember what's underneath it. Count up the money that's actually Angela's own: the $34,000 she'd already saved plus the $33,600 she adds over these 14 years comes to $67,600 of her own principal. The fee gap alone — $28,024 of pure cost difference — equals more than 40% of everything she personally put in. To make it visceral, look at just the first year on her current $34,000 balance: the 2.25% annuity skims $765 a year off her money, while the 0.10% index skims $34. That's $765 versus $34 — for the same investments, in the same account, doing the same job. The annuity costs her about twenty-two times as much, every single year, and the gap compounds.

And Angela is relatively lucky, because she's 14 years out, not 40. For a teacher with a full career ahead, the same fee gap compounds into life-changing money. A young teacher contributing $250 a month for 35 years at the same illustrative 6% ends with $348,076 in the 0.10% index versus $216,628 in the 2.25% annuity — a gap of $131,448. That is not a typo and not a rounding artifact: it's more than a third of the entire account, transferred out of a teacher's retirement and into fees, one quiet 2.25% slice at a time. The earlier in a career the annuity gets chosen, the larger the share of the final balance it consumes.

So that the lesson lands where it should and not an inch further: the villain here is specific. It is the high-fee variable annuity, sold by a commissioned non-fiduciary, placed inside an already-tax-deferred account. It is not every annuity — a low-cost annuity, or an immediate annuity bought at retirement for guaranteed lifetime income, can be a perfectly sound choice, as Lesson 30 will show. And it is not every 403(b) — Marcus's matched, low-cost plan is excellent, and so are countless university, governmental, and nonprofit plans. The problem is concentrated in one place: the K-12 voluntary multi-vendor menu, where a teacher with no match and no fiduciary is handed a dozen vendors and left to spot the one good door herself. Spotting it, and getting to it, is what the next section is for.

§2.3 — The escape: finding the low-cost door

If you've read this far and realized you're in a bad annuity — or you're staring at Angela's vendor list trying not to pick one — the situation is far more fixable than it feels, and the fix has a clear two-step shape. The crucial reassurance up front: you are not locked in for life, and stopping the damage costs nothing. The two steps are different jobs with different urgencies, so we'll take them in order, from Angela's own desk.

Step one is to stop the bleeding, and it's free and immediate: redirect your future contributions to the low-cost custodial option. Remember the salary reduction agreement from §2.1 — the form that tells payroll where your money goes. Angela signs a new SRA naming the 403(b)(7) custodial vendor (the tinted index option on her list), and starting with her very next paycheck, her $200 a month flows into the 0.10% door instead of the 2.25% one. Nothing about her existing balance has to move yet; no surrender charge is triggered, because she isn't withdrawing anything — she's only changing where new money lands. This single form is the highest-value action in this entire section, and it has no cost and no downside. Every future contribution immediately starts compounding at the better rate. The bleeding stops today.

Step two is the harder one, because it touches money that may be behind a surrender charge: whether to move the existing balance. This is a real calculation, not an automatic yes, and Angela should run it deliberately. Suppose her $34,000 sits in a variable annuity with a 5% surrender charge. Moving it now costs her a one-time $1,700. Leaving it costs her the ongoing fee drag — the 2.25% annuity versus a 0.10% index is a 2.15% difference — which on $34,000 is $731 every year, forever. Compare the one-time cost to the annual cost and the answer falls out:

The $1,700 surrender charge pays for itself in about 2.3 years of avoided drag — and Angela has 14 years to go, so escaping pays off many times over. The general lesson holds well beyond her exact numbers: a one-time surrender charge is usually cheap compared with years of ongoing 2%+ fees, which is why escaping a high-fee annuity almost always wins over a long horizon. The one thing to respect is the surrender clock — if you're a year from a chunk of money aging out of its surrender period, it can be worth waiting that year to avoid the charge entirely.

To do step two correctly, contact the annuity firm directly — not the agent who originally sold it to you. The selling agent earns nothing by helping you leave and may earn something by talking you out of it; the firm's service line can tell you, neutrally, your exact surrender penalty today and the date each contribution ages out of its surrender window. With those two facts you can decide whether to move now or wait. This is a fact-gathering call, not a negotiation — you're entitled to those numbers, and you need them before you act.

What if your plan offers no good vendor at all — no low-cost custodial option anywhere on the list? Three fallbacks, in order. First, ask the district to add one: benefits offices can and do add vendors when employees request them, and a low-cost custodian costs the district nothing to include. Second, if that stalls, fund a Roth IRA instead for your beyond-match retirement money — we cover the IRA fully in Lesson 18, so just hold the pointer for now. Third, if you also have access to a 457(b), that can serve as your low-cost escape route (more on the 457 later in this lesson). The point is that a bad vendor list never traps you completely; it just routes you to a different door.

Two plain resources make all of this easier, and they're worth writing down. The first is your district's vendor list itself, available from the benefits office — the same list on Angela's screen, which you're entitled to see and which spells out exactly which vendors your plan permits. The second is 403bwise.org, an independent nonprofit built specifically to help teachers navigate this exact problem; it rates vendors and maintains a district-by-district database so you can look up your own plan. Neither is a sales pitch. Both exist because this trap is common enough that a community grew up around escaping it — which is itself the reassurance to end on: you are very far from the first teacher to walk out of a bad annuity, and the path out is well-worn.

§2.4 — The catch-up only teachers get: the 15-year rule

There's one feature of the 403(b) that no 401(k), no 457, and no TSP has — a catch-up contribution available only to long-serving teachers, hospital workers, and church employees — and Angela, at 48 with a long tenure, is exactly who it's for. It's a smaller point than the annuity trap, but it's real money for the right person, so it's worth knowing precisely.

It's called the 15-year-service catch-up. The rule: if you've worked at least 15 years for the same qualifying employer — a school, hospital, home-health or health-and-welfare agency, or church — your plan may let you defer up to an extra $3,000 a year, on top of the normal $24,500 limit, with a lifetime cap of $15,000 of these extra deferrals. Two facts carry the weight here. The annual extra is capped at $3,000, and the lifetime total is capped at $15,000 — and unlike most retirement numbers, these two figures are fixed and never adjusted for inflation. So this is a $3,000-a-year door that stays open until you've used $15,000 of it across your career with that employer.

For Angela, this is a live possibility, not a hypothetical. She's 48 and approaching — or already past — 15 years in her San Antonio district. If her plan permits the 15-year catch-up, she could add up to $3,000 a year to her 403(b) beyond the standard limit, accelerating the supplement she's building on top of her TRS pension. That's a meaningful boost for someone 14 years from retirement who's trying to make up ground.

A few honest caveats, kept light because they matter less than the headline. The actual amount you can use each year is technically the least of three calculations — but for nearly everyone the binding numbers are the two you already know, the $3,000 annual and $15,000 lifetime caps, so don't over-engineer it. The 15 years must be with the same employer; changing districts resets the clock. And it's an optional plan feature — your plan document has to offer it, so the one action item is to ask the benefits office whether your plan permits the 15-year catch-up before counting on it.

One last sequencing note, because it stacks with another catch-up Angela's nearly eligible for: the age-50 catch-up ($8,000 in 2026) is a separate door she reaches at 50, and an eligible teacher can use both in the same year. When both apply, the IRS applies the 15-year amount first and the age-50 amount second. The practical upshot is simply that the two stack rather than compete — a long-serving teacher over 50 can layer the $3,000 15-year catch-up and the $8,000 age-50 catch-up on top of the standard limit, if the plan allows the first of them. For Angela, the move is small and concrete: ask whether her plan offers the 15-year catch-up, and if it does, it's extra tax-advantaged room she alone — by virtue of being a long-serving teacher — gets to use.

There is a second account that a large share of government and public-sector workers can use, sitting quietly alongside the 403(b) or the pension, and almost nobody outside government has heard of it: the 457(b). It deserves its own short stretch of this lesson because it isn't just another tax-deferred box that works like the ones we've already met — it carries a genuine superpower that no other retirement account in the entire system has, plus a separate contribution limit that lets the right person shelter roughly twice as much. We'll meet someone who can use both of those advantages, learn the one version of this account that carries a real and often-missed danger, and see honestly who this is actually for — because the headline features are powerful, and they only matter to a worker who has the income to reach them.

This is three beats. First, the two kinds of 457 and why one is meaningfully safer than the other (§3.1). Then the feature that makes the 457 unique — no early-withdrawal penalty — and how an early retiree uses it as a bridge (§3.2). And finally the double-dip: how having both a 403(b) and a 457 lets one person defer the full limit to each, and the special catch-up that goes with it (§3.3).

§3.1 — Two kinds of 457, and why one is safer

A 457(b) is a deferred-compensation plan — a name worth pausing on, because it tells you exactly what it does. You agree to defer some of your compensation: instead of taking that slice of your salary now, you have your employer set it aside into the account, untaxed, where it grows until you draw it out later. Mechanically, from where the saver sits, that feels identical to a 403(b) or a 401(k) — money leaves the paycheck before tax, lands in investments, and compounds. The 2026 elective-deferral limit is even the same $24,500 we've seen throughout this lesson. But underneath that familiar surface, the 457 splits into two genuinely different animals, and which one you have decides how safe your money actually is.

Meet Marisol Vega, because she has the good kind and she'll carry most of this section. Marisol is 46, a county public-health administrator earning about $72,000 a year. Her employer is a unit of local government, which makes her plan a governmental 457(b) — the version offered by state and local governments to their employees. Crucially, the law requires a governmental 457(b) to hold its assets in trust for the benefit of the employees. That phrase, held in trust, is the whole ballgame: it means Marisol's deferred salary is legally hers, walled off and protected from the county's own creditors, and it can be rolled over to an IRA or another plan if she ever leaves. Her money is as secure as the money in a 403(b) or a TSP. For Marisol, the 457 is simply an excellent, safe, second tax shelter.

The other kind is where the danger lives, and it catches a specific group of people off guard. A non-governmental 457(b) — sometimes called a tax-exempt 457(b) — is offered not by a government but by a private nonprofit: a hospital, a university, a charity. By law it can only be offered to a select group of management or highly compensated employees, which is why it's known as a top-hat plan — a plan for the people at the top of the org chart, not for rank-and-file staff. Picture a senior executive at a nonprofit hospital, well-paid, offered a top-hat 457 to defer part of a large salary. It sounds like a perk, and in many ways it is. But it carries a hazard the governmental version simply doesn't have.

Here is the hazard, stated plainly because it's the kind of thing that gets buried. In a non-governmental 457(b), the money you defer is not held in trust for you — it legally remains the employer's property. You have only the employer's unsecured promise to pay you later. That means if the nonprofit hits hard times — a bankruptcy, a ruinous lawsuit, an insolvency — your deferred compensation is exposed to the employer's general creditors, and you can stand in line with everyone else the organization owes money to. The retirement savings that felt locked away and safe can be reached by the hospital's creditors precisely because, in the eyes of the law, it was never separated from the hospital. And one more limitation compounds it: a non-governmental 457 generally cannot be rolled over to an IRA or another employer plan, so you can't simply move the money to safety when you leave. For a nonprofit executive, this is a real, frequently overlooked risk hiding inside a benefit that looks purely like an upside.

None of this is a reason for a nonprofit executive to refuse a top-hat 457 outright — for a high earner with no better shelter left, deferring at a high tax rate can still make sense, and many sponsoring organizations are financially rock-solid. The point is to know what you actually hold. The governmental 457(b), like Marisol's, is creditor-protected, held in trust, and rollable — a safe account in every sense that matters. The non-governmental top-hat 457(b) is a bet on your employer's continued solvency, and that bet belongs out in the open where you can weigh it, not hidden behind the comfortable assumption that all retirement money is untouchable. The two share a name and the same $24,500 limit; they do not share the same safety.

§3.2 — The feature no other account has: no early-withdrawal penalty

Now the superpower, and it's genuinely unique. Every other retirement account we've discussed — the 401(k), the 403(b), the TSP, and the IRA coming in the next lesson — punishes you for taking money out before age 59½ with a 10% early-withdrawal penalty stacked on top of the ordinary income tax you'd owe anyway. That penalty is the wall that keeps retirement money locked up until retirement. The governmental 457(b) is the one account that doesn't have that wall. Once you separate from service — once you leave the job — you can draw from a governmental 457(b) at any age with no 10% penalty at all. You still owe ordinary income tax on what you withdraw, exactly as you would in retirement; you simply skip the penalty that every other account would charge. Among the 401(k), the 403(b), the TSP, and the IRA, the governmental 457 stands alone in this.

This turns the 457 into something specific and valuable: an early-retirement bridge. Marisol plans to retire at 57, ten years before the standard 59½ penalty cutoff is even relevant and well before she'll claim other retirement income. The gap between leaving work and reaching 59½ is exactly the stretch where most people can't touch their retirement savings without the 10% bite — which is why early retirement is so hard to fund. Marisol's 457 solves precisely that. She can retire at 57 and draw from the 457 penalty-free to carry her across the two and a half years until she turns 59½, at which point all her other accounts open up penalty-free anyway. The 457 bridges the gap that would otherwise trap her money.

The dollars make the advantage concrete. Suppose Marisol withdraws $40,000 a year from her governmental 457 across that bridge, from 57 to 59½ — about two and a half years of living expenses. She owes ordinary income tax on that $40,000 either way; that's true in any account and isn't the point. The point is the penalty she avoids. In a 403(b), a TSP, or an IRA, that same $40,000 withdrawn before 59½ would carry a 10% penalty — $4,000 a year — purely for the crime of being early. In the governmental 457, that $4,000 a year simply doesn't apply. Over the roughly two and a half years of the bridge, that's about $10,000 in penalties she never pays — money that stays in her pocket instead of going to the IRS as a punishment for retiring on her own schedule.

There is a catch worth teaching carefully, because it's a gap that quietly undoes the whole advantage if you don't know about it — and it cuts in both directions, tracked by which dollars are which. The penalty-free treatment attaches only to money that originated in the governmental 457 itself. So if Marisol were to roll her 457 balance out into an IRA or a 401(k), those dollars would lose the penalty-free access the moment they landed — back behind the same 10% wall as everything else, until she's 59½. Rolling out forfeits the superpower. And it runs the other way too: if she rolled money into her 457 from a 401(k), a 403(b), or an IRA, that incoming money keeps its 10% exposure, and the plan tracks it in a separate bucket precisely so the penalty-free rule isn't accidentally extended to dollars that never earned it. The original 457 deferrals are penalty-free; rolled-in money is not.

That bidirectional rule is exactly why a near-early-retiree like Marisol may deliberately leave money in the 457 rather than consolidating it elsewhere. The instinct, drilled in by the earlier 401(k) lessons, is often to roll everything into one tidy account — and for most people most of the time, consolidation is the right move. But for someone counting on the no-penalty bridge between an early retirement and 59½, rolling the 457 out would trade away the single feature that makes the early retirement affordable. Knowing that the penalty-free access lives only inside the 457, and dies on the way out, is what stops a well-meaning rollover from quietly closing the bridge she was planning to walk across.

§3.3 — The double-dip: maxing a 403(b) AND a 457

Marisol's job comes with one more advantage, and it's the reason her county offers her two accounts instead of one: she has access to both a governmental 457(b) and a 403(b). That isn't redundant, and it isn't an accident — it unlocks something called the double-dip, which is worth defining precisely because it breaks a rule you might reasonably assume applies everywhere. Most retirement accounts share a single elective-deferral limit. If you had a 401(k) at one job and a 403(b) at another, you would not get to defer $24,500 to each; the two would share one combined $24,500 limit across both, and maxing one would eat into the other. The governmental 457 is the exception. Its deferral limit is entirely separate — not aggregated with the 403(b) or the 401(k) — so a person with both a 403(b) and a governmental 457 can defer the full $24,500 to each in the same year. That's the double-dip: two separate tax shelters, two full limits, $49,000 of deferrals in 2026 instead of $24,500.

What that second shelter is worth, for someone who can actually fill it, is striking, and it's the cleanest possible illustration of a separate limit. Picture Marisol with the cash flow to max her contributions — putting in the full $24,500 a year — across her 11-year runway from 46 to 57, at an illustrative 6% annual return (illustrative, as always in this lesson, never a promise). If she could only use one account, maxing that single shelter, those contributions grow to about $380,409 by 57. But because the 457 limit is fully separate from her 403(b) limit, she can run the same full $24,500 into both accounts — and the result is about $760,817. The second shelter doesn't add a little; it exactly doubles the outcome, an extra $380,409 of retirement savings, because the two limits genuinely don't touch each other. That clean doubling is the whole point of the double-dip made visible — and, just as plainly, it shows what the double-dip asks: a person with the income to put $49,000 a year away, which is precisely why the honest framing at the end of this section matters.

There's a second, lesser-known lever the 457 hands someone like Marisol as she approaches retirement: the final-3-years catch-up. This is a special provision — distinct from the age-50 catch-up we met earlier, and distinct from the 403(b)'s 15-year rule — that lets you supercharge a 457 in the home stretch. In each of the three calendar years before your plan's normal retirement age, you can defer up to twice the standard limit — $49,000 in 2026 instead of $24,500 — to the extent you have prior unused contribution room to make up for. It exists to let people who under-saved earlier in their careers pour money in at the end. But it comes with a firm restriction: in any year you use the final-3-years catch-up, you cannot also use the age-50 catch-up. They don't stack; you take whichever produces the higher number that year, not both.

One more difference belongs here, because it shapes how you should think about the 457 as an emergency resource: a 457's hardship rules are stricter than a 403(b)'s. Where a 403(b) hardship can be triggered by a fairly broad list of needs, a 457 requires a genuine unforeseeable emergency — a true, unexpected event beyond your control, claimable only after you've exhausted your other resources. The 457 is built to stay locked until you separate; it is not designed to be tapped for a planned or foreseeable expense. That's a feature, not a flaw, but it means you shouldn't count on the 457 as a flexible cash source the way the no-penalty-on-separation rule might tempt you to.

And now the honest part, because the double-dip can read like a flex if it isn't framed truthfully. Deferring $49,000 a year across two accounts — let alone pushing to $65,000 with the age-50 catch-up in both ($32,500 each) — is simply out of reach for most workers. The plain reality is that most people won't max even one retirement account, let alone two; the median saver is working to capture a match and build an emergency fund, not searching for a second tax shelter. The double-dip isn't advice for everyone. It's a tool that becomes valuable only once the earlier, higher-priority steps are handled — and where the 457 sits against your other goals belongs to the priority waterfall, which we take up in its own lesson, not here. For the specific person who has both accounts and the income to fill them — a Marisol with strong cash flow and an early-retirement plan — the double-dip is a quietly enormous advantage. For everyone else, it's good to know the account exists and what it can do, and to come back to it when the foundation underneath it is built.

We've spent the lesson watching what happens when a tax-advantaged account gets stuffed full of high-fee products by someone who profits from selling them. Now meet the opposite — the account built as if someone actually sat down and asked, "what would the cheapest, simplest, most honest retirement plan in the country look like?" It's the Thrift Savings Plan, the TSP, and it's the retirement account for federal civilian employees and members of the military. It is, by most fair measures, the single best large retirement plan in America, and it's worth holding up as a benchmark precisely because it proves a public-sector defined-contribution plan can be best-in-class rather than predatory.

What makes it the gold standard is almost boringly simple: it charges next to nothing. The all-in cost of holding money in a TSP fund runs around 0.035% a year — about thirty-five cents per $1,000 invested, annually — which is among the very lowest expense ratios anywhere in the world of investing. That one number is the whole reason this account is special, and it's why a federal worker or service member who simply leaves their money in the TSP and contributes enough to get the match is, almost by default, doing better than most savers anywhere. Tomás Rivera — 38, an Army Sergeant First Class with about eighteen years of service, stationed at Fort Liberty, married to Carla, roughly two years from a twenty-year retirement — has $89,000 sitting in his TSP right now, and the next three beats are about why that account is quietly one of the best things he owns: the funds and the fees that make it so cheap (§4.1), the government money that lands on top of his own (§4.2), and the one Roth wrinkle unique to the TSP (§4.3).

§4.1 — The funds and the fees: why 0.035% changes everything

The screen above is Tomás's TSP fund allocation — the whole investment menu of the plan, which is the first surprise, because the entire menu fits on one page. Where a 403(b) teacher might face forty annuity vendors and a 401(k) saver a sprawling list of funds, the TSP gives you a deliberately tiny set of choices: five core funds, plus a family of one-decision target-date funds built from those five. That smallness is a feature, not a limitation — fewer ways to overpay, fewer ways to get lost. Let's walk it the way Tomás would read it, naming each fund as we go, because each is just a plain-English container for one slice of the market.

The G Fund is the safe corner: it holds special U.S. government securities issued only to the TSP, and it has a property no other fund here has — its share value never drops in nominal terms. The dollar amount you put in cannot fall; only the interest rate it pays changes. That rate is reset monthly from the average yield on longer-dated Treasury securities, and as of June 2026 it's 4.500% — though because it resets every month, that exact figure is the sort of number to check live rather than memorize. For someone close to needing the money, the G Fund is the place that simply won't lose principal.

The F Fund is the bond fund: it tracks a broad U.S. bond index, so it holds a wide swath of the American bond market in one fund. Unlike the G Fund, its value can rise and fall with bond prices — it's the diversifier that behaves like the bond sleeve of any normal portfolio. The C Fund is the workhorse stock fund: it tracks the S&P 500, meaning it holds the 500 largest U.S. companies, so a dollar in the C Fund buys a tiny slice of corporate America's biggest names. The S Fund is its companion — a small- and mid-cap completion index, which is the technical way of saying it owns the rest of the U.S. stock market that the C Fund leaves out, the smaller and medium-sized companies. Hold C and S together and you own essentially every public U.S. company. And the I Fund is the international stock fund: it holds companies based outside the United States, the slice that lets a portfolio participate in the rest of the world's economy rather than betting on America alone.

If reading even those five and deciding how to split between them feels like one decision too many, the TSP has the same answer Lesson 16 gave for the 401(k): the Lifecycle funds, labeled the L Funds. An L Fund is the TSP's target-date fund — a single fund built entirely from the five core funds above, mixed for someone retiring around a particular year, and automatically rebalanced over time along a glide path that shifts from stock-heavy when you're young toward safer holdings as the target year approaches. You pick the one whose year roughly matches when you'll need the money, and it does the diversifying and the de-risking for you, exactly the way Maya's target-date fund did in Lesson 16. The crucial detail unique to the TSP: the L Funds add no extra fee layer. You pay only the blended cost of the underlying core funds, so the one-decision path here is just as cheap as building it yourself — which is not true in many other plans, where the convenient option quietly costs more.

Now the number that makes all of this matter. The 2025 net expense ratios on the core funds run from about 0.034% to 0.059% a year — the G Fund at 0.034%, the C Fund at 0.035%, the I Fund at 0.054%, the S Fund at 0.059%. To put 0.035% in human terms, it's thirty-five cents a year for every $1,000 you hold. Compare that to the variable annuity from the front of this lesson skimming 2.25% — $22.50 per $1,000 — and you're looking at a fund that costs roughly one sixty-fourth as much. The TSP itself reports that fewer than 1% of the roughly 170,000 funds tracked across the industry charge less than its average fund does. This is, almost literally, as cheap as investing gets.

Here's what that fee edge is worth to Tomás specifically, using his current $89,000, left alone for twenty years at an illustrative 6% a year with no new contributions — a figure meant to show the mechanism, never a promise of returns. Kept in the TSP at 0.035%, that $89,000 grows to about $283,556. Rolled into a typical advisor-managed IRA charging 1.00% a year — the kind of rollover a salesperson might pitch him the week he separates — the same $89,000 over the same twenty years at the same 6% grows to only about $236,143. The difference, $47,413, is money the TSP simply keeps in Tomás's account instead of handing to a middleman, built from nothing but the gap in fees. Same money, same market, same time — one number lower on the cost line, $47,413 more at the end.

One honest caution so the picture is complete. The TSP does offer an optional mutual-fund window — a doorway to thousands of outside funds beyond the five core ones — but it carries extra fees the core menu doesn't: roughly $55 a year in administrative cost, a $95 annual maintenance fee, and about $28.75 per trade, with a $10,000 minimum to use it. That's not the default and it's not where most people belong; it exists for the investor with a specific need the core funds can't meet. For nearly everyone, including Tomás, the five funds and the L Funds are the whole answer, and the rock-bottom fee is exactly why.

§4.2 — The government's 5%: the match and the vesting

The screen above is where Tomás sets his TSP contribution — for military members it lives inside myPay, the Defense Department's pay portal, and you elect a percentage of your basic pay rather than a flat dollar amount. This is the screen where the free money is captured or forfeited, and it works the same way for two big groups of people, so it's worth naming both. FERS is the Federal Employees Retirement System, the retirement framework for federal civilian workers. BRS is the Blended Retirement System, the corresponding framework for service members like Tomás, which blends a (smaller) military pension with a TSP that the government contributes to. We treat Tomás as a BRS member precisely because he gets that government match — and the match mechanics below are identical for FERS civilians and BRS military.

Here is the government's offer, and it's the same generous structure Lesson 16 taught for the employer match, just with the federal government on the other side. First, an automatic 1% of your basic pay goes into your TSP whether or not you contribute a cent of your own — free, unconditional. Then, on top of that, the government matches the first 5% you contribute: dollar-for-dollar on the first 3% you put in, and fifty cents per dollar on the next 2%. Work it through and the rule is clean — contribute 5% of your basic pay, and the government adds 5% of your basic pay (the 1% automatic plus a 4% match). Contributing more than 5% earns you no additional government money; contributing less leaves part of that 5% on the table. So the single most important instruction here is the same free-money rule from Lesson 16: always contribute at least 5%.

Watch it land on Tomás. His basic pay is $4,350 a month. He contributes 5% of that, which is $217.50 a month out of his own pay. In return, the government puts in its automatic 1% — $43.50 — plus its match, which works out to another $174 (dollar-for-dollar on the first 3% is $130.50, fifty cents on the next 2% is $43.50). Add the government's pieces together and it's $217.50 a month — exactly matching his own $217.50. That's $2,610 a year of government money he gets simply for contributing his 5%, money that vanishes entirely if he contributes nothing. And because it's invested in that ultra-cheap TSP and left to compound, the government's 5% alone — $217.50 a month for fifteen years at an illustrative 6% — grows to about $63,000. That $63,000 is the price tag on the sentence "I didn't bother to contribute": it's not a small perk forgone, it's tens of thousands of dollars of real retirement wealth that only exists if he reaches for it.

Now vesting, which is the rule for when this money is irrevocably yours — the same idea from Lesson 16, that your own contributions are immediately yours but the employer's portion may carry a waiting period. In the TSP the good news is broad: every dollar you contribute yourself, and every dollar of the government's matching, is yours immediately — you keep all of it the moment it lands, no matter when you leave. The only piece with a waiting period is the automatic 1%. For military and BRS members like Tomás, that 1% vests after two years of service; for FERS civilians it's three years. Tomás, eighteen years in, is long past both thresholds, so for him every dollar — his, the match, and the automatic 1% — is fully his. The vesting rule mainly matters to someone in their first couple of years deciding whether to stay.

Two real traps in the fine print, both worth flagging because they catch service members who think they're doing everything right. First, only basic pay counts toward the match. Military compensation includes allowances like BAH (housing) and BAS (food) that are non-taxable — and because they're non-taxable, they're TSP-ineligible and don't enter the match math at all. A service member who mentally counts those allowances as part of "my pay" can misjudge what 5% really needs to be; the percentage is on basic pay only. Second, the timing differs by who you are. For military members the government's 4% match doesn't begin until you complete two years of service — the automatic 1% starts early, around the sixty-day mark, but the matching portion waits. Federal civilians, by contrast, get the full matching from day one. None of this changes the headline instruction — contribute at least 5% — but a brand-new service member should know the match is still warming up, while a civilian's is live immediately.

§4.3 — Traditional or Roth TSP

The pre-tax-versus-Roth decision got its full treatment in Lesson 16 — the clean question of whether your tax rate is likely to be lower or higher in retirement than it is today, with pre-tax winning if you expect a lower rate later and Roth winning if you expect a higher one. That whole framework carries over to the TSP unchanged, so there's no need to re-walk it here; the TSP offers both a traditional balance and a Roth balance, and you choose between them by the same logic. Tomás currently has all $89,000 in the traditional side — he's never used the Roth TSP — which is a perfectly common starting point, not a mistake to panic over.

There is, though, one wrinkle genuinely unique to the TSP, and it's important to understand so it doesn't surprise you: all of the government's contributions go into the traditional balance, even if you direct your own contributions to Roth. The automatic 1% and the entire match are always pre-tax money, no matter what you choose for yourself. So a service member who contributes to the Roth TSP ends up with two buckets automatically — their own Roth contributions on one side, and all the government's traditional money on the other. That's not a problem; it's just the plan's mechanics, and it means even a committed Roth contributor will always have some traditional balance growing alongside.

As for which side fits Tomás, the gentle, evenhanded nudge leans toward giving the Roth TSP a look — not a mandate, just a fit worth considering. A relatively young service member is often in a comparatively low tax bracket now, and a good deal of military pay can be tax-advantaged in ways civilian pay isn't (combat-zone pay, for instance, can come in tax-free). Paying little or no tax on money going in, and then drawing it out tax-free decades later, is exactly the situation where Roth tends to win under the Lesson 16 logic. That's a reason for Tomás to consider routing some future contributions to the Roth side, while keeping in mind he'll always have the government's traditional money too — a natural split that gives him tax flexibility later. It's a consideration, weighed by his own read of his bracket now versus in retirement, not a rule.

One last item carries over from Lesson 16 §6.2: the high-earner Roth catch-up mandate applies here too. If your prior-year FICA wages exceeded the indexed threshold (roughly $150,000), any age-50-or-older catch-up contributions must be made as Roth rather than pre-tax — the same 2026 rule, the same reach. It doesn't touch Tomás at his current pay, and it doesn't touch the large majority of federal and military savers, but it's the same provision in force across all these accounts, so it's worth knowing it lives here as well.

There is one moment in a service member's financial life when more money is in play, and more strangers want a piece of it, than at any other: the moment of separation. The day Tomás Rivera — the 38-year-old Army Sergeant First Class at Fort Liberty, with $89,000 sitting in his Thrift Savings Plan — finishes his last day in uniform, that $89,000 stops being a quiet payroll deduction and becomes a liquid, movable balance that anyone with a financial product to sell would love to relocate. He is roughly two years from a 20-year retirement, so this isn't his moment yet. But it's close enough that the pitches have already started arriving, and it's worth meeting the decision now, calmly, before he's standing inside it.

Separating service members are peak targets for rollover pitches for a simple structural reason: the TSP is one of the cheapest investment accounts on Earth, which means almost any place a salesperson can move your money to pays them more than leaving it where it is. That conflict doesn't make every pitch a scam, but it does mean the burden of proof sits on whoever wants you to move. This section takes the decision in two parts: the four things you can actually do with the account, and why 'leave it in' is usually the right one (§5.1); then the income gap that tempts a costly cash-out, and the larger retirement context Tomás is weighing as a Blended Retirement System member (§5.2).

§5.1 — The four options, and why "leave it in" is usually right

When you separate from the service, your TSP doesn't disappear and it doesn't automatically follow you anywhere — it sits exactly where it is until you decide. And there are exactly four things you can do with it, which map almost one-for-one onto the four 401(k) job-change options from Lesson 17, just with the TSP in the starring role. Three of them keep your money working and tax-sheltered; one of them is the costly mistake. Knowing what each actually does, before a salesperson explains it for you, is the whole point of this beat.

Option A is to leave it in the TSP. As long as your vested balance is at least $200, you can keep the account for as long as you want after you separate. You keep its rock-bottom fees, you keep full investment control — you can still move money between the G, F, C, S, and I funds and the Lifecycle funds whenever you like — and you keep every withdrawal option the plan offers. The only thing you give up is the ability to make new contributions from a paycheck you no longer receive. For most separating members, this is the right default, and the reason is the fees, which we'll quantify in a moment.

Option B is to roll it into a new employer's plan. If Tomás takes a civilian job with a 401(k) or 403(b) that accepts incoming rollovers, he can move the $89,000 there, untaxed, and keep it compounding. The appeal is consolidation — one account to watch instead of two — which is real, but it only makes sense if the new plan's fees and fund menu are at least as good as the TSP's, and very few are.

Option C is to roll it into an IRA he opens and controls. Also untaxed, also still compounding, and it gives him the widest possible investment selection — but an IRA's funds usually cost more than the TSP's, and as we'll see, that difference is exactly where the money quietly leaks out. This is the option the rollover pitches almost always steer toward, precisely because it's the one that can carry a 1%-a-year advisor fee.

Option D is to cash out — take the $89,000 as a check. This is the costly mistake, and the math is the same brutal math from the 401(k) lessons: the withdrawal is taxed as ordinary income, and because Tomás is under 59½, it gets hit with a 10% early-withdrawal penalty on top. A balance that large could lose a third or more to tax and penalty the instant it lands, and that's before counting the decades of growth it would have produced if left alone. Cashing out a six-figure retirement account at 38 isn't spending $89,000; it's spending the several hundred thousand it would have become. We'll come back to what tempts people into it in §5.2.

Now the low-fee case for leaving it in, made concrete, because this is the argument that quietly wins. The TSP's funds cost about 0.035% a year — roughly 35 cents per $1,000 invested, among the very lowest costs available anywhere. A typical advisor-managed IRA might charge around 1.00% a year, all in. That gap looks like a rounding error and is anything but. Run Tomás's $89,000 forward 20 years at an illustrative 6% gross return, with no new contributions, and the difference is stark:

Leaving the money in the TSP rather than moving it to that 1% IRA is worth $47,413 to Tomás over those 20 years — and even against a cheaper 0.50% IRA, the TSP keeps about $23,876 more. That return figure is illustrative, never a promise; markets don't deliver a smooth 6%. But the fee gap driving the difference is not illustrative at all — it's a fixed, guaranteed drag that applies in good markets and bad. Nothing in the advisor IRA changes what the money is invested in enough to justify skimming that much off it every year. The $47,413 is, almost entirely, just the cost of moving the money somewhere more expensive.

Which is exactly why you should be careful about who's doing the urging. FINRA — the brokerage industry's own regulator — warns plainly that financial professionals who recommend rolling your TSP out into an IRA often earn commissions or ongoing fees for doing so, while leaving your money in the TSP earns them little or nothing. That's not an accusation that every advisor is acting in bad faith; it's a description of a built-in conflict of interest you have to manage. The defense is simple and non-negotiable: before you move a dollar, get every fee — the advisor's and the funds' — compared against the TSP's, in writing, as a percentage and in dollars. A pitch that can't or won't put the total cost on paper next to the TSP's 0.035% is telling you something. And there's a one-way-door element that raises the stakes: once you roll the TSP fully out and the account hits zero, you generally can't undo it — the account closes and the door back in shuts behind you.

There's also a feature you'd quietly forfeit by rolling out, and it has a name worth learning: the Rule of 55. The Rule of 55 is an IRS provision that says if you separate from service in or after the calendar year you turn 55, you can take withdrawals from that employer's plan — including the TSP — without the 10% early-withdrawal penalty, even though you're under 59½. (Income tax still applies; the rule waives only the penalty.) For certain public-safety roles the threshold is even earlier, age 50. Here's the catch that ties it back to the rollover decision: the Rule of 55 belongs to the employer plan, not to you personally — so if you roll the TSP into an IRA, you lose it, and an IRA's own penalty-free age snaps back to 59½. A member who separates between 55 and 59½ and might need to tap the money therefore has a concrete, dollars-and-cents reason to keep it in the TSP rather than roll it out. Tomás, separating around 40, is too young for this particular gap to bite him yet — but it's exactly the kind of feature a rollover pitch won't mention, because losing it costs you and not them.

One more thing the pitches rarely volunteer: the door swings the other way too. You can roll other accounts into the TSP — an old 401(k), a 403(b), a governmental 457(b), or a traditional IRA can all be moved in, even after you've separated, as long as your TSP account is still open. (A Roth IRA can't be rolled in, and Roth employer money goes to the Roth side of the TSP.) So for someone who values the TSP's fees, consolidation can run toward the TSP rather than away from it — gathering scattered old accounts into the cheapest plan available instead of paying someone to scatter them into a pricier one.

Whichever direction you move money, do it the way Lesson 17 taught: as a direct rollover, never an indirect one. A direct rollover sends the money institution-to-institution — you never touch it, nothing is withheld, nothing is taxed. An indirect rollover routes the check to you first to redeposit within 60 days, with a chunk withheld up front, and if you miss the window the whole thing converts into a taxed, penalized cash-out. There is never a good reason to run retirement money through your own hands. The boring direct rollover is the safe one, every time.

So four options, and three of them are fine. Leave it in for the unbeatable fees and the features you'd otherwise lose; roll it to a new plan only if that plan genuinely beats the TSP; roll it to an IRA with eyes open to the cost; and don't cash out. For the large majority of separating members, the boring first option — leave it in — is the right one, and the burden of proof is on anyone urging otherwise to show you, in writing, that their account beats 0.035%.

§5.2 — The transition gap, and BRS vs the old pension

There's a specific, predictable squeeze that drives separating service members toward that cash-out button, and it deserves naming because it's a real cash-flow problem, not a character flaw. Call it the transition gap: the stretch between your last military paycheck and the moment your next reliable income — retirement pay, VA disability compensation, or a new civilian salary — actually starts arriving. None of those switch on the instant you separate. Retirement and VA payments can take weeks or months to begin and catch up, a new job may not start the day after the old one ends, and the bills don't pause to wait. Staring at a gap with no paycheck coming in, a $89,000 balance sitting right there starts to look less like a retirement account and more like an emergency fund.

It isn't one, and the move is to make sure you have an actual emergency fund so the TSP never has to pretend to be one. This is exactly the role of the cash cushion from the earlier emergency-fund lesson: money parked safely outside the market specifically to cover a stretch like this without touching retirement savings. Tomás and Carla have built roughly $18,000 in savings — enough to carry several months of expenses across a transition — which is precisely what lets him treat the TSP cash-out as the non-option it should be. The discipline is to fund the gap from the cushion, not the retirement account, because the cushion costs you a little forgone interest while a TSP cash-out at 38 costs you the 10% penalty, the income tax, and the several hundred thousand dollars that $89,000 would have compounded into. The emergency fund exists so the worst financial moment of the transition never forces the worst financial decision.

One narrow but important note while we're on loans: if Tomás has an outstanding TSP loan when he separates, he needs to pay it off within 90 days. An unpaid TSP loan balance left past that window is treated as a taxable distribution — and if he's under 59½, a penalized one — so a loan he was comfortably repaying from a paycheck can quietly convert into a tax bill the moment the paychecks stop. It's a small box to check at separation, and an expensive one to forget.

All of this sits inside a larger question Tomás is genuinely weighing two years out: how his retirement actually works, because he's covered by the Blended Retirement System rather than the older pension-only setup. It's worth understanding the trade evenhandedly, because the headline — 'BRS gives you a smaller pension' — is true and also misleading on its own. The Blended Retirement System, which covers everyone who joined on or after January 1, 2018 (and those who opted in during the 2018 window), is exactly what its name says: it blends two pieces. It pays a smaller defined-benefit pension than the old 'legacy High-3' system — the pension multiplier is 2.0% per year of service under BRS versus 2.5% under legacy, which works out to a pension roughly 20% smaller at the 20-year mark. In exchange, it adds the thing the legacy system never offered: a portable, fast-vesting TSP match (the government money Tomás has been collecting all along, which we covered earlier), plus Continuation Pay — a mid-career cash bonus paid to those who agree to serve on.

That trade is not strictly worse, and which side wins depends entirely on how long someone serves. The legacy pension only pays out at all if you reach the 20-year cliff; serve 19 years and 11 months under the old system and you walk away with no pension whatsoever. BRS, by contrast, hands you vested TSP matching money you keep no matter when you leave, which is why it's the better deal for the large majority of service members — most of whom separate before the 20-year mark and would have gotten nothing from the legacy pension. For the career service member who does stay 20-plus years, the richer legacy multiplier can come out ahead. Tomás, about two years from his own 20-year retirement, is unusual in that he's one of the people approaching the cliff — but the point here is to place him correctly as a BRS member weighing a transition, not to pin down his exact future pension.

And we'll deliberately stop short of that number. The precise pension a 20-year career produces — the dollar figure, the percentage of base pay, how the multiplier and his High-3 average combine — is its own detailed calculation, and it gets its full treatment in Lesson 56. For this lesson, what matters is the TSP decision in front of him: he's a BRS member with $89,000 in the cheapest account available, a transition gap to bridge with savings rather than a withdrawal, and a rollover landscape full of people who'd profit from moving his money somewhere more expensive. Get those three things right — keep the TSP, fund the gap from the cushion, and don't be talked out of either — and the harder pension arithmetic can wait for the lesson built to handle it.

We've now met all three siblings one at a time. This block puts them in a single frame, because the most useful thing about the 403(b), the governmental 457(b), and the TSP is not how each works in isolation — that you already have from Lesson 16's 401(k) machinery, which all three inherit — but where they quietly differ. The same $24,500 deferral mechanic runs through every one of them, so the table below deliberately holds that column constant and lets the differences do the talking: who can use each account, whether free employer money is on the table, what the funds inside typically cost, whether the 10% early-withdrawal penalty applies, and which special catch-up belongs to which account. Read it as a map of the differences, not a re-teaching of the basics.

Three rows in that table are where real money is won or lost, so each is worth a sentence of plain meaning. The employer-match row is the reason Tomás's TSP and Marcus's matched 403(b) outrank the accounts with no match — free money is the highest guaranteed return there is, exactly as the waterfall lesson framed it. The typical-fees row is the whole annuity story compressed into one cell: the 403(b)'s ~2%+ figure is the high-fee variable annuity, the ~0.10% figure is the 403(b)(7) custodial index escape, and the TSP's ~0.035% is the benchmark every other account is judged against. And the early-withdrawal row is the governmental 457(b)'s signature gift — no 10% penalty on your own deferrals after you separate, at any age, which none of the other three offer.

The bottom row — special catch-ups — needs its own table, because each account carries a different extra-savings rule on top of the ordinary age-based catch-ups, and they're easy to confuse. The cheat sheet below lays out every catch-up these accounts offer in 2026, who gets it, and how much it adds.

One line in that table corrects a misconception almost everyone holds, so it's worth stating flat out: the age 60–63 super catch-up of $11,250 is used INSTEAD OF the regular $8,000 catch-up, not stacked on top of it. A 61-year-old does not get $8,000 plus $11,250 — she gets $11,250, full stop, for a total of $35,750. And it's a four-year window only: at age 64 the super catch-up disappears and she reverts to the regular $8,000. People routinely overstate their 60–63 limit by exactly $8,000 because they add the two; the rule is to take the larger of the two, which is the $11,250.

There's a second confusion the table can quietly invite, so let's separate two numbers that get conflated constantly. The $24,500 is the elective-deferral cap — the limit on what you choose to send in from your own paycheck. The 415(c) cap of $72,000 is a different ceiling entirely: it's the total of everything that lands in your account in a year, your deferrals plus all employer money plus any after-tax contributions, before catch-ups. They are not the same wall. A federal worker contributing $24,500 with the government adding its 5% is nowhere near the $72,000 ceiling; the $72,000 cap only becomes the binding constraint in plans that allow large additional employer or after-tax contributions. When a plan document mentions both numbers, it isn't a contradiction — they're measuring two different things.

Where do these accounts sit in the order you fund them? That's the priority-waterfall question, and the full answer lives in Lesson 11 — but the shape of it follows directly from this table. The match comes first, always, because no fee or fund choice beats a guaranteed return on free money; that's why Tomás funds his TSP to 5% and Marcus funds his 403(b) to the match before anything else. Then the cheaper container wins the next dollar. And that's the one forward-point worth planting here: for dollars beyond the match, a low-cost IRA (Lesson 18) usually outranks a high-fee, no-match 403(b), for exactly the reason the 403(b)(7) escape exists — when there's no match to chase, you route money to wherever the funds are cheapest. The IRA is the topic of its own lesson (Lesson 18), so we'll leave its mechanics there; the point for now is simply that it belongs in the conversation the moment a 403(b) charges annuity-level fees and offers no match to offset them.

Across this lesson the same family of accounts met very different people, and the right move was different for each one. Here they are together, so you can find the situation closest to yours and see what it actually asks of you.

Angela — escape the annuity, and the gap is the prize. At 48, the San Antonio teacher has $34,000 sitting in a high-fee variable annuity inside her 403(b), adding $200 a month, with no employer match and no Texas state income tax. Her single highest-value move is to stop the bleeding: open a 403(b)(7) custodial account — the mutual-fund version of a 403(b) — with a low-cost index vendor, and redirect her future contributions there via a new salary-reduction agreement, which is free and immediate. The reward for switching is concrete. Run her $34,000 plus $200 a month to her TRS retirement age of 62 — fourteen years — at an illustrative 6% gross: the annuity's ~2.25% all-in fee nets her about $101,529, while the ~0.10% index account nets about $129,554. That roughly $28,000 gap is money she keeps simply by changing the container — and not by saving a dollar more, because it works on her own money she already has in play: the $34,000 sitting there today plus the $33,600 she adds at $200 a month over those fourteen years, $67,600 in all, just routed into a cheaper wrapper. Her lesson: a 403(b) is a fine tax shelter; the villain is the product chosen inside it.

Marcus — the good 403(b), so capture the match, then check the fees. At 41, the Chicago history teacher has the version of this account that works: his district matches 3%, so contributing 6% of his $68,000 means he puts in $4,080 and the district adds $2,040, for $6,120 a year — and because his employer contributes, his plan is ERISA-covered, with the stronger protections that brings. He's doing the most important thing right, and his move is to confirm that rather than assume it, then look one level deeper at the expense ratios of the funds his contributions actually buy. A correct 6% sitting in needlessly pricey funds is the next dollar he can save. His lesson, and the lesson he holds for the whole cast: a match changes everything — not every 403(b) is the trap, and his is proof.

Tomás — the TSP gold standard, so contribute 5% and leave it there. At 38, the Army Sergeant First Class earns $4,350 a month in base pay and is about two years from a 20-year retirement. As a Blended Retirement System member he gets a match, and the arithmetic is the cleanest in the lesson: he contributes 5% of base pay — $217.50 a month — and the government adds an automatic 1% plus a 4% match, another $217.50, so a 5% contribution earns 5% from the government and yields 10% of pay going in. Contributing less than 5% forfeits part of that free money; the government's 5% alone, $217.50 a month for fifteen years at an illustrative 6%, is about $63,000 he'd simply walk away from. Inside the TSP he can hold a single L (Lifecycle) fund or the cheap core funds, all at roughly 0.035% — lower than almost any fund anywhere. And when he separates, the move is usually to leave his $89,000 right where it is: at the TSP's ~0.035%, that balance grows to about $283,556 over 20 years at an illustrative 6%, versus about $236,143 in a 1.00% advisor IRA — the TSP keeps roughly $47,000 more, which is why the burden of proof is on any rollover to beat it. His lesson: the TSP is the benchmark; don't let anyone talk you out of it without showing you the fees in writing.

Marisol — two accounts, so capture the match, build the bridge, and double-dip. Marisol Vega is the county worker from §3, worth restating plainly: she's 46, a county public-health administrator earning about $72,000, and her employer offers her both a governmental 457(b) and a 403(b) — and she plans to retire early, at 57. Her situation rewards three distinct moves. First, capture any match her plan offers, for the usual reason. Second, use the 457's no-penalty bridge for her early retirement: governmental 457 money she contributed can be withdrawn after she separates with no 10% penalty, even before 59½ — so if she pulls $40,000 a year from 57 to 59½, she avoids the $4,000-a-year penalty a 403(b) or TSP would charge, about $10,000 over the bridge (ordinary income tax still applies either way). Third, if her budget allows, double-dip: because the 457's deferral limit is fully separate from the 403(b)'s, she can defer $24,500 to each, $49,000 total. Run that for her eleven years to 57 at an illustrative 6% and maxing both produces about $760,817 versus about $380,409 from maxing just one — the second tax shelter exactly doubles the result. Her lesson: when you have both accounts, they're separate opportunities, not a choice between.

If none of these is exactly you, you're somewhere among them — and the through-line holds regardless of which sibling account you hold. Find out whether there's a match and capture it first; find out what the funds inside actually cost and escape to a cheaper container if they're expensive (a 403(b)(7) index account, or an IRA for beyond-match dollars); know which special features are yours — the 457's penalty-free bridge, the 403(b)'s 15-year catch-up, the TSP's rock-bottom fees; and treat every annuity pitch in the teachers' lounge as the sales call it is until you've verified the person and compared the fees. The account type on your statement is just the door you came through. What matters is the same in every one: capture the free money, pay the lowest fee you can, and let compounding do the rest.

Every account in this lesson is a quiet pile of money, and where there's a pile of money there's someone whose job is to get paid moving it. But the frauds that circle a 403(b) have a particular flavor, and it's worth naming precisely, because the danger here usually doesn't look like a scam at all. It looks like a helpful person at a folding table in the break room, with a clipboard and a warm manner and the word "district" somewhere in their pitch. The whole skill of this section is learning to see the machinery behind that friendliness — and, just as important, learning that none of what follows is your fault to catch unaided, because the system is built to make the salesperson look official. Here's what to watch for, and exactly where to take it if something feels off.

The teachers'-lounge problem

The signature 403(b) fraud isn't a crude one — it's a commissioned insurance agent or a captive company rep who shows up where teachers and nonprofit staff actually are: the teachers' lounge, the on-campus open-enrollment table, the new-hire benefits meeting. They're selling high-fee variable or fixed-indexed annuities — the very products §2's annuity trap is built around — and the move that makes it work is the way they present themselves. They appear on the district's vendor list, so they describe themselves as "district-approved" or a "district-appointed rep," and the phrase does enormous quiet work. It sounds like the district vetted them. In the K-12 voluntary 403(b) market, that list is usually open-access: almost any insurer can get onto it, and the district reviewed neither the fees nor the suitability of what's being sold. Angela, our San Antonio teacher, met exactly this person — a friendly rep at a lunchtime table who called himself the district's 403(b) representative — and the reassurance worth saying plainly is that being approached this way, and even believing the "approved" framing, is not naïveté. It's the predictable result of a system that lets a salesperson borrow the district's authority without earning it.

The "free lunch" seminar

A close cousin arrives as an invitation rather than a table: a free-meal seminar, or an "educational workshop" held on campus during open enrollment, framed as a chance to learn about your retirement options. Treat the free lunch as what it is — the price of admission to a sales presentation. This isn't cynicism; it's documented. When the SEC, FINRA, and state regulators examined firms running these events, 100% of the seminars billed as "educational" were actually sales presentations, about half used exaggerated or misleading advertising, and a slice appeared to be outright fraudulent. The products pushed are most often high-commission annuities with surrender periods of seven to fifteen years. So the rule for one of these is simple and protective: go if you like, eat the lunch, take the materials home — but never sign anything during the event. The pressure to commit before you leave the room is the tell, and "let me think about it" is a complete and acceptable sentence.

The dual-license tell

Here's a single check that quietly exposes a great many of these sellers, and it turns on a licensing detail most people never learn. A variable annuity is a securities product, so anyone selling one must hold both a state insurance license and a FINRA securities registration (a Series 6 or 7, registered through a broker-dealer). A fixed or fixed-indexed annuity, by contrast, needs only a state insurance license. The consequence is a clean tell: an "advisor" who is insurance-licensed only will not appear in FINRA's BrokerCheck at all. So if the person across the table calls themselves an advisor, talks like one, and yet can't be found in BrokerCheck, that gap is real information — they're very likely insurance-only, legally able to sell you fixed and indexed annuities but not securities, and not registered the way a securities professional would be. It doesn't by itself prove bad intent. It tells you which database to verify them in, and it tells you the "advisor" label may be doing more work than the license behind it.

Before you trust anyone with this money, verify them — it's free, it takes a few minutes, and the verification stack here has one extra step most people skip, which is exactly the step that catches the teachers'-lounge seller.

Check a securities professional in FINRA's BrokerCheck (brokercheck.finra.org) and in the SEC's tools at Investor.gov / IAPD (adviserinfo.sec.gov), which cross-link and show licensing, employment history, and — read this part — any disclosure events and regulatory actions. Verification is not endorsement; appearing in the database only confirms registration, so the disclosure section is the part that actually tells you something.

Then the step almost everyone misses: verify an insurance or annuity agent separately, through your state's Department of Insurance or the NAIC's free agent lookup (sbs.naic.org). This is the whole point of the dual-license tell — because the annuity seller is often insurance-licensed only, they won't be in BrokerCheck, and the only place to confirm their license, the lines they're authorized to sell, and any regulatory actions is the state insurance system. Run both checks, not just one.

And know where the recourse actually lives, because the intuitive answer is wrong for these plans. Governmental 403(b) and 457(b) plans — public-school, public-college, state and local government plans — are exempt from ERISA, which means the Department of Labor's EBSA cannot help you here and you can't sue under ERISA the way a 401(k) participant could. For an abusive annuity or insurance agent, the real recourse is your state Department of Insurance (via the NAIC), or your state securities regulator and the SEC for a securities product or registered rep — not the federal labor department, where many victims waste months complaining to the wrong agency.

To report fraud or a bad-business practice of any kind, the FTC at ReportFraud.ftc.gov — you can report even if you didn't lose a dollar, and the report feeds a database that roughly 2,800 law-enforcement agencies use to build cases. And if your TSP account is compromised or you suspect identity theft on it, call the ThriftLine at 1-877-968-3778 and lock the account.

The most important line, the same one the regulators themselves lead with: if something feels wrong, don't let embarrassment stop you from reporting it. The worry that you'll be judged for not handling your own money perfectly is precisely the feeling these salespeople count on, because shame keeps people quiet and keeps the pitch working on the next teacher down the hall. You don't deserve that shame, and reporting protects the next person at least as much as it protects you.

If you read the annuity section and felt your stomach drop — because the product it describes is the one you've been paying into for years, the variable annuity a friendly rep signed you up for in the teachers' lounge or at an open-enrollment table back when you started — this part is for you, and it carries no lecture. You are not the cautionary tale. You're the far more common case: a teacher or nonprofit worker who did the responsible thing, started saving, and got steered into the expensive default because the expensive default was the only person standing in the room.

First, the thing that matters most: this is not a failure of your intelligence, and it is not a thing you should have caught. The system was built to produce exactly this outcome. A 403(b) began life in 1958 as an annuity-only account, the insurance industry got there first and never left, and your district most likely handed you a vendor list it never vetted for fees — then let a commissioned salesperson, who is not a fiduciary and owes you only a weak 'suitability' standard, set up shop where you work and present the costly product as the normal one. Being caught by that is not on you any more than it would be your fault for trusting a person your employer appeared to endorse. The salesperson's whole job was to make the annuity feel like the obvious choice, and they're good at it. The regret you might feel reading the fee math is real; the self-blame underneath it isn't earned.

Second, and this is the part that actually changes things: the money you've already paid in fees is in the past and can't be recovered, but almost everything that matters about this account is still ahead of you. What you control is the years to come — and you can change the trajectory of those years starting with your very next paycheck, for free, without touching a dollar of your existing balance. So the move here isn't to fix everything at once. It's to take the single highest-value step first and let the rest follow.

Here's what 'better from here' looks like, in order.

Redirect your future contributions now — this is the high-value first step

You do not have to move your existing balance to stop the bleeding. You can file a new Salary Reduction Agreement — the one-page form that tells payroll where your 403(b) money goes — directing every future dollar to a low-cost option instead of the annuity. If your district's vendor list includes a 403(b)(7) custodial account (the kind that holds plain index mutual funds rather than an insurance contract) from a low-cost provider, point your contributions there. This is free, it's immediate, and it stops new money from entering the high-fee product today. Everything else on this list can wait; this one shouldn't. Redirecting the next contribution is the move that captures most of the value, and it's the one that takes five minutes.

Get the real numbers on the annuity you hold

Before you decide what to do with the existing balance, you need two facts the sales agent has no incentive to volunteer: the all-in annual fee and the surrender schedule. Call the insurance company directly — the firm, not the agent who sold it to you — and ask in writing for the total annual cost (mortality-and-expense charge plus sub-account fees plus any rider) and the surrender-charge schedule, including the exact dates each contribution becomes free to move. A typical variable annuity runs somewhere around 2.25% all in; on Angela's $34,000 balance, a 2.25% drag is about $765 a year skimmed off the top, against roughly $34 a year in a 0.10% index account. You want those numbers in hand and in writing before you weigh anything.

Weigh moving the existing balance against the surrender charge

Existing money is the harder call, because annuities lock it with a surrender charge — a one-time penalty, often around 5% and declining over several years, for moving the balance out early. The instinct is that the penalty makes escape not worth it. The math usually says otherwise. On Angela's $34,000, a 5% surrender charge is a one-time $1,700. But staying in the annuity costs her the ongoing fee drag — the difference between roughly 2.25% and 0.10%, about 2.15% a year, which on her balance is about $731 every single year, forever. Pay the $1,700 once and you recover it through lower fees in about 2.3 years; after that, the savings are pure. The one-time surrender charge is real, but the ongoing drag almost always outweighs it — so escaping a high-fee annuity tends to pay for itself fast. The only caveat is to mind the surrender clock: if a chunk of your balance ages out of its penalty in a few months, it can be worth waiting that short stretch to move that piece for free.

If the sale was misrepresented, report it for the next teacher

If you were told this was a 'district-approved' or 'recommended' product when it wasn't, or the costs and surrender lock-up were never disclosed, that's worth reporting — not because it's likely to undo your situation, but because it protects the next person who gets pitched in the same lounge. For an annuity or insurance agent, the right venue is your state's Department of Insurance, reachable through the NAIC consumer page, not the federal Department of Labor (most public-school plans are ERISA-exempt, so the DOL can't help here). If the seller was a registered rep handling a variable product, you can also file with the SEC and FINRA. Reporting isn't on you to fix the system, but it's how the next teacher avoids the lounge table.

You don't have to carry this as a verdict on your competence, and you don't have to fix it all in one sitting. The fees already paid are spent; the part of the story that decides how this account does from here is the part you're still writing. Redirect the next contribution to a low-cost option — that's the step that matters most — and the rest can follow at whatever pace fits your life. That's not just consolation; it's genuinely where almost all the leverage is.

The move

It's the first week of school, or open-enrollment season, and a friendly rep is at a table in the lounge. They catch Angela between classes with a genuinely helpful-sounding offer: "Let me help you get your 403(b) set up — I'll walk you through the paperwork, it'll take ten minutes, and you'll be saving for retirement by your next paycheck." It sounds like a favor. It removes a chore she's been meaning to handle and feels vaguely guilty about. And that's exactly why it works. Here's the machinery underneath the warmth.

What's actually being proposed

"Set up your 403(b)" sounds like opening a generic account. What's actually being proposed is enrolling Angela in one specific product the rep happens to sell: a variable annuity inside her 403(b). The account type and the product inside it are two different decisions, and the pitch quietly fuses them — "set up your 403(b)" becomes "sign this annuity contract," as if there were no other way to do it. There is; the menu almost always contains a low-cost custodial option too. But the rep isn't paid to mention that one.

What's in it for them

Follow the incentive, because it's the whole story. The rep earns a sales commission the moment Angela signs the annuity, and then an ongoing cut every year after — the stacked fees from §2, the mortality-and-expense charge plus sub-account fees plus any rider, running roughly 2.25% all-in. On her $34,000 balance, that's about $765 a year skimmed off the top, versus about $34 a year in a comparable low-cost index option. The rep isn't necessarily lying about anything; they're just not volunteering that the help costs Angela hundreds of dollars a year, every year, forever. And this connects to the fiduciary idea from the earlier advisor lesson: a commissioned insurance agent is not a fiduciary — they're held only to a weaker "suitability" standard, not a legal duty to put Angela's interest first — which is precisely the difference between this person and a fee-only fiduciary advisor who is bound to her interest and isn't paid by the product they recommend.

The redundancy decode

There's a specific piece of nonsense buried in this pitch that's worth being able to name out loud, because it's the technical heart of why the annuity doesn't belong here. An annuity's headline selling point is tax deferral — your money grows without being taxed each year. But a 403(b) is already tax-deferred; that's the entire point of the account. So wrapping a tax-deferred annuity inside a tax-deferred 403(b) buys Angela a tax shelter she already has, and charges her the extra annuity fees for the privilege. As the people who teach this put it bluntly: there's no reason to put a tax-deferred annuity inside a tax-deferred account. The deferral is redundant; only the cost is new.

Legit vs. not — the spectrum

This is not a story where every annuity is evil and every rep is a crook, and saying so honestly is what makes the real warning land. If a low-cost provider or a fee-only fiduciary helps Angela enroll in a 403(b)(7) custodial account holding cheap index funds, that's genuinely good help — the account type is a perfectly fine tax shelter, and many governmental, university, and nonprofit 403(b) plans are excellent (the colleague down the hall whose plan has a low-cost index menu and an employer match is in a genuinely good account). Even some annuities have a place: a fixed or immediate annuity at retirement, or a low-cost variable annuity at roughly 1% or less all-in, can be reasonable for the right person. The villain is narrow and specific: the high-fee variable annuity sold inside an already-tax-deferred account by someone whose pay depends on selling it. The problem isn't the 403(b), and it isn't the word "annuity" — it's that exact product in that exact place from that exact seller.

The DIY substitute

Here's the reassuring part the pitch obscures: the thing the rep is offering to "handle" is something Angela can do herself, for free, in about the same ten minutes. She can ask her benefits office directly for the low-cost 403(b)(7) custodial option — the mutual-fund version of the account, not the annuity contract — and enroll in that. If her district's list has no good 403(b) vendor, the 457 or, for someone like Tomás, the TSP can serve the same purpose at rock-bottom cost. The enrollment paperwork is the benefits office's job to walk her through, and they have no product to sell her. The rep's real value-add here isn't access — Angela already has access — it's just the appearance of having someone do it for her, priced at hundreds of dollars a year.

The questions that expose it

Angela doesn't have to diagnose the rep's character. She just has to ask four plain things a sales move struggles to answer cleanly — and listen to whether the answers come back clear or vague.

"Are you a fiduciary, in writing, legally required to act in my best interest?" (A fee-only fiduciary says yes plainly and puts it on paper; a commissioned agent will dodge toward "I always do what's right for my clients," which is not the same thing.)

"Is what you're recommending an annuity, or a custodial mutual-fund account — a 403(b)(7)?" (This forces the fused pitch back apart into its two real decisions.)

"What is the total annual cost — every fee combined — as a percentage and in actual dollars on my balance?" (Vagueness here is the tell; a 2%-plus answer next to a low-cost option's 0.1% answers itself.)

"Is there a surrender charge, and for how many years is my money locked in?" (A yes — typically around 5% for five to seven-plus years, with each new contribution often restarting its own clock — means the product is built to trap the money, the strongest reason to walk.)

The decode, in one line: "Let me help you set up your 403(b)" can mean a free favor you could do yourself in ten minutes, or it can mean let me sell you a high-fee annuity you don't need, wrapped in a tax shelter you already have. The four questions — especially the cost in dollars and the word "annuity" — separate the two faster than reading the rep's friendliness ever could. Slow down, ask them, and let the answers, not the warmth, decide.

If this lesson left you with a low hum of anxiety — that your accounts are second-class because they're not called a 401(k), that the annuity trap means the whole system is rigged against people like you, or that doing this right requires becoming the kind of person who reads fund prospectuses for fun — it's worth taking a moment to set that weight down, because the real picture is far kinder than the worry suggests.

Start with the accounts themselves, because that's where the doubt usually sits. The 403(b), the governmental 457(b), and the TSP are not lesser cousins of the 401(k) — they're siblings. They use the same paycheck-deferral mechanics, share the same 2026 base limit of $24,500, and offer the same age-50 catch-up. A public servant's retirement accounts are as good as any private-sector worker's, and in the case of the TSP they are better than almost anything a private employer offers: its funds cost in the neighborhood of three-and-a-half to six cents per hundred dollars a year, lower than the overwhelming majority of funds in existence. Tomás isn't behind because he's in the military; on his $89,000, staying in the TSP's roughly 0.035% instead of rolling to a 1% IRA is worth, on an illustrative 6% path over twenty years, about $47,000 more. The structure you have is genuinely good. The annuity trap was never about the account being bad — it was about one expensive product sold inside an otherwise fine shelter, and Marcus's matched, ERISA-covered 403(b) is proof the same account type can be excellent.

Then the fear that getting this right takes expertise you don't have. It doesn't. The decision that matters most is short, and this lesson already handed it to you: capture any match that exists, and pick the low-cost option on your menu. That's the part that counts. For Tomás, capturing the match means contributing at least 5% so the government adds its full 5% — about $2,610 a year of free money he'd forfeit entirely at 0%. For Marcus, it's the 6% that earns his district's 3%. And picking the low-cost option is almost always a single line: the 403(b)(7) custodial index fund where your district offers one, or an L fund (the TSP's self-managing target-date option) if you're in the TSP. Capture the match, choose the cheap index, sidestep the annuity, and you've done the part that actually moves the needle. You don't have to become an expert. You have to read the menu and step around one product — both of which this lesson walked you through.

And if you're worried that a wrong choice locks you in, it almost never does. The form that controls where your 403(b) money goes — the Salary Reduction Agreement — can be changed; you can redirect future contributions to a better option any time you decide to, for free. If you're auto-enrolled at a rate or a fund you didn't choose, you can move it. Almost nothing here is permanent, which means there's no single high-stakes decision you have to get perfectly right on the first try — only a menu to read and a default to improve whenever you're ready.

Your accounts are good, the most important move is simple, and almost everything is changeable. Read the menu — capture any match, pick the low-cost option — and sidestep the annuity, and let compounding do the rest. That's enough — and it's well within what you can do, starting now.

Why is my 403(b) full of annuities — is that normal?

Unfortunately yes, it's normal for a K-12 teacher's plan, and it's also the single biggest thing worth fixing. The 403(b) started life in 1958 as an annuity-only account — mutual funds weren't even allowed inside one until 1974 — so the insurance industry got there first and never left. As recently as 2014, about 76% of all 403(b) money was sitting in annuities. The product that dominates the menu is the variable annuity, an insurance contract whose all-in fees stack up: a mortality-and-expense (M&E) charge of roughly 1.20%, sub-account fund fees of about 0.90%, and a small rider of about 0.15%, totaling roughly 2.25% a year. On Angela Morales's $34,000 balance, that 2.25% skims about $765 every year, versus about $34 a year in a 0.10% custodial index fund — the same money, about twenty-two times the cost. Here's the part that makes it indefensible: an annuity's main selling point is tax deferral, but a 403(b) is already tax-deferred, so wrapping an annuity inside it duplicates a shelter you already have and charges you extra for the privilege. The damage over Angela's 14 years to her TRS retirement age is real — at an illustrative 6% gross, her annuity grows to about $101,529 while the same contributions in a low-cost 403(b)(7) index account reach about $129,554, a gap of about $28,024 (these are illustrative figures, not promises). None of this means annuities are evil or your 403(b) is a bad account — the account is a perfectly fine tax shelter, and fixed or low-cost annuities have their place at retirement. The villain is specifically the high-fee variable annuity sold inside an already-tax-deferred account by a commissioned salesperson who isn't your fiduciary. The fix is to ask your benefits office for the district's vendor list and request the 403(b)(7) custodial option — a mutual-fund account from a low-cost provider — and redirect your future contributions there.

403(b) vs 457 — can I have both, and should I?

If you're a state or local government worker, you can often have both, and having both is one of the most powerful and least-known moves in public-sector saving. It's called the double-dip, and it works because the 403(b) and the governmental 457(b) have separate contribution limits that don't share or aggregate. In 2026 the elective-deferral limit is $24,500 per account, so someone with access to both can defer $24,500 to the 403(b) and another $24,500 to the 457(b) — $49,000 total — versus the $24,500 a single account allows. (Contrast that with a 401(k) and a 403(b), which share one combined limit — those two can't both be maxed.) Take Marisol Vega, a 46-year-old county public-health administrator earning about $72,000 who has access to both a 457(b) and a 403(b) and plans to retire at 57. Over her 11 years to 57 at an illustrative 6% gross, maxing one account ($24,500/yr) grows to about $380,409; maxing both ($49,000/yr) grows to about $760,817 — the second tax-shelter exactly doubles the result, because its limit is genuinely separate (illustrative, not a promise). The 457 carries a second gift on top of the extra room: a governmental 457(b) has no 10% early-withdrawal penalty after you separate from service, at any age. For Marisol, retiring at 57 — before the usual 59½ threshold — that means she can pull, say, $40,000 a year from the 457 to bridge the gap to 59½ and skip the $4,000-a-year penalty a 403(b) or IRA would charge, about $10,000 saved over a 2.5-year bridge (ordinary income tax still applies either way). So should you do both? If you can afford to fund both, the 457 is usually the more valuable second account precisely because of that no-penalty access. If you can only fund one fully, weigh whether you'll need money before 59½ — if you might retire early, the 457's penalty-free bridge tilts it toward the 457.

Is the TSP really as good as everyone says?

For once, the hype is earned — the Thrift Savings Plan is genuinely the gold-standard low-cost retirement plan, and it's worth understanding why so you can hold it up as the benchmark every other account should be measured against. The headline is the fee. The TSP's individual funds run roughly 0.034% to 0.059% a year (the C Fund tracking the S&P 500 is about 0.035%), figures so low that as of January 2026 fewer than 1% of the roughly 170,000 funds tracked by FactSet reported lower expenses. Put that next to the 1%-plus an advisor-sold IRA often charges, or the ~2.25% of a variable annuity, and the gap compounds into real money. Take Tomás Rivera, a 38-year-old Army Sergeant First Class with about $89,000 in his TSP: left in the TSP at 0.035% and growing at an illustrative 6% gross for 20 years with no new contributions, it reaches about $283,556; the same money rolled into a 1.00% IRA reaches about $236,143 — the TSP's low fee keeps about $47,413 more in his pocket (illustrative, not a promise). The menu is deliberately simple and excellent: five core funds — G (government securities, which never loses nominal value), F (a broad bond index), C (the S&P 500), S (smaller U.S. companies), and I (international) — plus Lifecycle (L) target-date funds built entirely from those five, auto-rebalanced quarterly, with no extra fee layer. The G Fund is unusual and worth knowing: its share value never drops, only its interest rate changes, currently about 4.500% (June 2026, reset monthly). One honest caveat to the 'best account' story: the TSP's optional mutual fund window lets you reach outside funds but layers on extra fees — about $55 admin plus $95 maintenance a year plus $28.75 per trade, with a $10,000 minimum and a 25% cap — so it's worth it only if you genuinely need funds beyond the stellar core menu. For almost everyone, the core funds are the whole answer.

What happens to my TSP when I leave the military or federal service?

You have four options, and the most important thing to know in advance is that three of them keep your money working and one of them quietly costs you a fortune — and that separating service members are a peak target for people who want to talk you into the wrong one. The four: (1) leave it in the TSP, (2) roll it to a new employer's plan, (3) roll it to a traditional or Roth IRA, or (4) cash it out. Cashing out is the trap — under 59½ it triggers ordinary income tax plus a 10% early-withdrawal penalty, and you forfeit everything the money would have become. For Tomás Rivera, about two years from a 20-year retirement with $89,000 in his TSP, the strongest default is usually to leave it right where it is. You can keep a TSP account forever with a vested balance of $200 or more, you keep its roughly 0.035% fees and full investment control, and you simply stop making new employee contributions. Why not just roll it to an IRA? Two reasons. First, fees: rolled into a 1.00% IRA that same $89,000 reaches about $236,143 over 20 years at an illustrative 6% versus about $283,556 left in the TSP — about $47,413 more kept by staying (and even a 0.50% IRA leaves the TSP ahead by about $23,876). Second, a feature you lose the moment you roll out: the Rule of 55, which lets you take penalty-free TSP withdrawals if you separate in or after the year you turn 55 (age 50, or 25 years of service, for many public-safety roles) — gone if the money moves to an IRA, where the wait is 59½. Here's the scam to expect: FINRA warns that advisors earn commissions for moving you out of the TSP but little or nothing for leaving you in, so an unsolicited 'roll your TSP into our IRA' pitch is almost always more expensive for you. The TSP itself will never contact you about investment opportunities. One more thing to plan for: the cash gap between your last military paycheck and your retirement or new-job income is exactly what tempts a costly cash-out — fill it with an emergency fund, not your TSP. You can roll other accounts (a 401(k), 403(b), governmental 457(b), or traditional IRA) into the TSP, but a total cash-out zeros the account and ends that ability.

I'm a teacher with no employer match — is the 403(b) even worth it?

Yes — just not at the top of your priority order, and only if you get into the right version of it. Angela Morales is exactly this case: a 48-year-old San Antonio teacher earning $58,000 with a $34,000 403(b) balance to which she adds $200 a month, and no employer match. The match is the one feature that makes an account a must-fund-first priority, so without it, the urgency drops — your emergency fund, any high-interest debt, and an IRA (which we cover in Lesson 18) generally come ahead of extra 403(b) money, and the full priority order is the waterfall lesson, Lesson 11. But 'lower priority' is not 'skip it.' The 403(b) still does two genuinely valuable things even with no match: it lowers your taxable income today (Angela's $200 a month goes in pre-tax), and it compounds untaxed for years. What actually decides whether it's worth it for Angela is the fee, not the match. If her $200 a month sits in a 2.25% variable annuity, the account is fighting her — over her 14 years to TRS retirement age at an illustrative 6% gross it grows to about $101,529, while the identical contributions in a 0.10% custodial index 403(b)(7) reach about $129,554. That's a roughly $28,024 gap created by nothing but fees, on money she worked for — about $67,600 of her own is in that account by the end (her $34,000 starting balance plus the $33,600 she adds over those years). So the honest answer is: the no-match 403(b) is worth using once the foundation is in place and an IRA is funded — but only if you steer your contributions into a low-cost 403(b)(7) custodial account instead of a high-fee annuity. Get the cheap version, and a no-match 403(b) is a solid supplement to her TRS pension. Get stuck in the expensive one, and you'd often be better off in an IRA you open yourself.

Can I have a 403(b)/457/TSP and an IRA?

Yes — these are completely separate buckets with their own limits, and an IRA is a smart complement to any of them, not an either/or choice. Your workplace account (a 403(b), a governmental 457(b), or the TSP) has its own 2026 elective-deferral limit of $24,500, and the IRA has its own, much smaller limit on top of that, so funding one does nothing to reduce what you can put in the other. They're different tools: the workplace account lets you defer far more and sometimes comes with a match or, in the TSP's case, rock-bottom fees, while an IRA you open yourself typically offers a wider menu and the cheapest index funds available — which is exactly why an IRA is often the right home for dollars beyond a match, especially when your workplace menu is expensive. We teach the IRA in full in Lesson 18 (its limits, the Roth-versus-traditional income rules, and where it sits in the order), and the full priority sequence between all these accounts is the waterfall in Lesson 11, so we won't re-derive either here. The one-line version for now: having a 403(b), 457, or TSP never disqualifies you from also funding an IRA — for many public-sector savers, using both is the plan. For someone like Marisol Vega, who already has a 403(b) and a 457(b), an IRA is a potential third bucket on top of those two separate workplace limits; for Angela, stuck in a high-fee annuity 403(b), a low-cost IRA is one of the cleaner escape hatches for her beyond-essentials savings.

Should I choose Roth or traditional in my 403(b)/TSP?

It's the same decision you'd make in a 401(k), and it comes down to one question: is your tax rate likely to be lower or higher in retirement than it is right now? A traditional contribution goes in pre-tax, lowers your taxable income this year, and is taxed when you withdraw it later — a bet that your rate then will be lower. A Roth contribution is the mirror image: you pay the tax now and withdraw it completely tax-free in retirement — a bet that your rate later will be higher (or simply a hedge against not knowing). Lower rate later favors traditional; higher rate later favors Roth; genuinely unsure favors splitting between the two. Roth versions exist across all three account types — Roth 403(b), Roth 457(b) where the plan offers it, and Roth TSP. A couple of specifics worth carrying. For a federal or military saver, every dollar of government match in the TSP lands in the traditional balance even if your own contributions go to Roth TSP — the match can't be Roth, so a Roth-contributing service member naturally ends up with both types, which is itself a reasonable hedge. And the 2026 high-earner rule from Lesson 16 applies here too: if your prior-year FICA wages topped roughly $150,000, your age-50-and-up catch-up contributions must be Roth — a change to plan around, not to fear, and one that doesn't touch the large majority of teachers and government workers below that threshold. We worked the full Roth-versus-traditional math in Lesson 16's §6, so this is just the public-sector footnote: the logic is identical, the Roth option is available, and the only twists are that the TSP match is always traditional and the high-earner catch-up may be forced to Roth.

This is the L20 interactive, and it puts the lesson's central choice in your own hands — the difference a fee makes, run on your numbers instead of a character's. Enter a starting balance, a monthly contribution, the years until you'll need the money, and a fee, then toggle between the annuity rate (about 2.25% all-in, the high-fee variable annuity) and the custodial index rate (about 0.10%, the 403(b)(7) low-cost option), and watch the career gap open up live between the two. The whole annuity-trap section collapses into one question this answers for your exact situation: how much is the fee inside my account actually costing me over the years I have left? A second tab compares the accounts themselves — flip between a 403(b), a governmental 457(b), and the TSP to see how their limits, fees, and access rules differ, including the double-dip: max one account and you defer $24,500, max a 403(b) and a 457 together and you defer $49,000, and the tool shows that the second tax-shelter exactly doubles the result because its limit is genuinely separate. Every figure recalculates from your inputs using the same illustrative 6% gross and the same formulas worked through this lesson, and the defaults reproduce the lesson's canonical figures exactly — Angela's roughly $28,024 annuity-versus-index career gap on her $34,000 balance and $200 a month over 14 years, and Marisol's double-dip doubling from about $380,409 to about $760,817. Every number is illustrative, never a promise. It runs entirely in your browser with useState only — nothing is stored, nothing is sent anywhere; close the tab and your numbers are gone.