# Barista, Coast, side income: work you choose FIRE's founding taboo died of ridicule. It held that if you still work, you are not really retired. But the corpus shows that almost every FIRE retiree keeps producing, and that many of them make money without going looking for it ([[voices-from-real-retirees]]). The research shows that work in **small doses** is one of the best predictors of well-being after a career: retirees in "bridge employment", the chosen halfway house between career and full retirement, show better physical and mental health than comparable full stops (Zhan, Wang, Liu and Shultz, 2009; [[meaning-and-identity]]). And the arithmetic shows that a few thousand dollars of income in the early years buy more plan safety than hundreds of thousands of dollars of extra capital ([[sequence-of-returns]]). This chapter builds the case for **work you choose**. It lays out the forms it takes (Barista, Coast, organic side income, the targeted temporary return, and the forced return, which deserves a straight look). It puts numbers on the power of it, including everything that rides along with a small earned income: health coverage, Social Security credits, and a softened sequence. It also tells the unpleasant truth about **decay**, because employability melts, and the option is worth most when you need it least. It is an asset, and it has to be maintained. It closes with the toolbox for work in small doses. The chapter's thesis fits in one sentence. The ability to earn a little, for a long time, whenever you choose, is the best defensive asset in this book, better than gold, better than TIPS. It is also the only one that makes your life richer at the same time as your plan. ::: cle A change of status In the plan, future earnings from work are not **counted** (the plan has to hold without them, [[how-much-you-need]]). But they are **cultivated**, and that distinction is the whole chapter. They are margin, not a parameter. And margin has to be maintained (the network, the skill, the paperwork ready to go) so that it is there on the day the sequence, the itch or the accident calls for it. A well-built FIRE is not "never work again". It is "never **have to** work again". And the difference is worth half a point of withdrawal rate. ::: ## The forms, from chosen to forced **Barista FIRE.** Capital covers 50% to 80% of your spending and chosen, dosed work covers the rest, continuously. It is the most powerful form financially, because it shortens the uncovered phase, the bridge years before any pension starts, and cushions it. It is also the most demanding on the choice of work. The risk of the model is the "fun" activity that turns back into an obligation at twenty hours a week. The accounts recommend work with a light chain of command, and work you can quit without drama. **Coast FIRE.** The capital is enough **eventually**: it will coast to a conventional retirement on compounding alone. You stop saving and work only to cover current spending. It is the gentlest form. The job continues but the pressure switches off, and the accounts describe how sharply the office changes once saying no becomes possible again. It is also the slowest, because full freedom waits. **Organic side income.** This is the majority case in the corpus, and it is not a plan. It is income that simply **arrives**: the project that starts paying, the gig somebody offers, the blog, the expertise people come asking for. You take it without counting it, and you dose it, with your hand always on the switch. **The targeted temporary return.** This is the pure anti-sequence tool: one to three years of work through a bad stretch of market, or after a shock in your life (flexibility's "form number one", [[flexibility-in-practice]]). It has one condition, employability kept alive, and that is the next section. **And the forced return.** This is the plan failing, and it is rare in the corpus. Survivorship bias is real, but plans built to modern standards break rarely ([[voices-from-real-retirees]]). The big retirement surveys confirm it outside FIRE too: about one American retiree in four goes back to some work, and that "unretirement" is most often anticipated and chosen, not distress (Maestas, 2010). When it does happen, the accounts show it seldom looks like the nightmare of starting over at a job you hated. It looks more like a Barista you did not pick: part-time, temporary, hard on the ego but livable. Knowing that defuses part of the fear that feeds one more year ([[one-more-year]]). ## The numbers: why this is the best defensive asset in the book Here is what earlier chapters established piece by piece, pulled together. **The sequence effect.** $12,000 a year of income through the first eight years of a plan that spends $40,000 cuts net withdrawals by 30% across the fragile window. In the standard simulations (variant C of the founding example, [[sequence-of-returns]]), failure falls from about 18% to about 8%. Only another $200,000 to $300,000 of capital would match that. The hour you work inside the fragile window is the best-paid hour of your financial life. **And everything that rides along.** The same small income does more than land in the checking account. - **The income itself**, the part you can see. - **Health coverage, the big one.** A part-time job that carries benefits takes the largest single line out of an early retiree's budget, and it is worth asking about before you ask about the pay. Without it, earned income cuts both ways, because marketplace subsidies key off your modified adjusted gross income: a paycheck moves that number, and it can shrink the credit as easily as a change of job can remove the question. Run it for your own state and your own year before you assume either direction ([[us-healthcare-and-social-security]]). - **Social Security credits.** Covered earnings buy quarters of coverage, four a year at most, and the dollar amount behind one credit is indexed every January, so read the current one off ssa.gov rather than off any book. The bigger mechanism sits in the benefit formula: it averages your highest 35 years of indexed earnings, so anyone who left before working 35 full years has zeros in that average. Every year you bill in early retirement replaces one of them, and the check waiting at 67 goes up for good. - **A softened sequence**, the effect above, the largest of the four by far, and the reason the other three are a bonus rather than the point. Add them up and a $10,000 job during the bridge years is worth more to the plan than $10,000 of avoided withdrawals, sometimes a good deal more, depending on what it does to your premium and to your benefit. No asset in the portfolio comes near that ratio. **And the psychological effect**, unquantifiable and unanimous. Retirees who **know** they can earn again sleep better and spend better (permission, [[the-psychology-of-spending]]). They go through crashes as spectators rather than hostages. ::: attention The unpleasant truth: the option melts Employability is an asset with a steep **decay** rate. Technical skills go stale (three to five years in software, more slowly elsewhere). Networks evaporate, because contacts move on and references age. And age discrimination is real, measured in the field by Neumark, Burn and Button (2019), more than 40,000 fictitious applications and callback rates that drop sharply with age. Landing a conventional salaried job after 50 or 55, outside a rare specialty, is hard, and the accounts confirm it. The planning consequence is clean. The "go back to work" option is worth **a great deal** in the first ten years of a plan, and less and less after that. That is exactly the shape of the melting buffer ([[refilling-the-buffer]]). It is a **first-decade** margin, never something to invoke to reassure a 70-year plan. Like any asset, it holds its value only if you **maintain** it. The veterans who kept theirs describe the same minimum upkeep. - A practice that keeps the skill visible: the one gig a year, the open source contribution, the course, the article. - A network you feed: the two lunches a quarter that keep you present in a profession. - The paperwork **ready**: the entity opened, even dormant, the invoicing set up, so you can bill inside a week. Two to four full-time-equivalent weeks a year. That is the insurance premium on the best defensive asset in the book, to be paid knowingly, or let lapse knowingly. Just not by default. ::: ## The toolbox for work in small doses The American market is unusually well equipped for work in small doses, largely because so much of it is already contract work. Five vehicles cover almost every case. **Negotiated part-time where you already work** is the underrated one. Dropping to a four-day week, or to three days, over the last two years is the partial version of one more year ([[one-more-year]]), and it comes with the thing that is hardest to buy on your own: employer health coverage, as long as you stay above your plan's hours threshold. Ask what that threshold is before you propose anything else. This is the canonical soft exit. **W-2 part-time work elsewhere** is the low-friction option. Somebody else withholds the taxes, the Social Security credits accrue on their own, and quitting is a two-week conversation. Ask about benefits eligibility first: a schedule that clears the threshold is worth more than a higher hourly rate that does not. **1099 contracting and consulting** pays the most per hour and gives you the finest control over the dose, which is why it is the default for a FIRE retiree with a professional skill. It costs you the employer, in both senses. You buy your own coverage, and you owe self-employment tax, both halves of Social Security and Medicare instead of the half a paycheck quietly hides, with part of it deductible against income tax. Price the work net of all that, never gross. **Seasonal and project work** suits people who want the year concentrated: a busy season, then months genuinely off. Tax season, the parks, the harvest, events, a term of teaching. The tradeoff is that the busy season stops being negotiable once you say yes. **And volunteering with a real role** serves the retirees whose usefulness does not need an invoice ([[meaning-and-identity]]). | Vehicle | Best at | Health coverage | Friction to start | |---|---|---|---| | Part-time where you work now | The soft exit | Yes, above the hours threshold | A conversation | | W-2 part-time elsewhere | Steady income, hands off | Sometimes, ask first | An application | | 1099 contracting | The highest rate, the finest dose | No, and self-employment tax on top | An entity and an invoice | | Seasonal or project work | A concentrated year | Rarely | An application, every season | | Volunteering with a role | Usefulness without billing | No | A commitment | The trap sits in the third row. Contracting looks like the best-paid line right up until you subtract your own premium and your own payroll tax, and then a benefits-eligible part-time job at half the hourly rate often wins outright. Do that subtraction before you choose. Which vehicle you pick matters less than having one ready to go. The paperwork is close to free (an EIN in ten minutes, a separate bank account, a liability policy where the trade needs one), so what decays is never the filing. It is the client list and the reference who will still take your call. ## The design choice: what place in your plan? Three stances hold together. Pick one on purpose ([[building-your-plan]]). **Full FIRE.** No income planned and none cultivated: the plan carries everything. It calls for the prudent rate and the full margins of this book, and for accepting that the option will melt. **Structural semi-FIRE** (Barista or Coast). Income is a **parameter** of the plan, sized and simulated, and the side income controls exist for exactly this ([[using-the-fire-simulator]]). It is the most efficient plan on paper, and the most demanding on the quality of the work. **FIRE with the option maintained.** This is the default stance the book recommends. Nothing is counted, everything is cultivated. The plan stands alone, employability is maintained as a first-decade margin, and organic side income is welcomed when it turns up. The consistency test is simple. Your withdrawal rate, your floor and your employability upkeep have to tell the **same** story. The 4.2% plan justified by "I can always go back to work", run by someone who has let their employability rot, is the most common inconsistency in the movement. ::: exemple Sam's bridge, the work-you-choose version Sam leaves at 47, with a plan that stands alone at 3.7%. He makes one design decision, FIRE with the option maintained, and gives it a real base. He targets about $15,000 a year of consulting, billed through an entity he opened **before** he quit (six weeks a year, on his terms), which nets out to roughly $10,000 in the household budget once self-employment tax and costs are paid. Every year of the bridge, the ledger reads the same. - The $10,000 brings withdrawals down to $40,000, a current rate of 3.0%. - His health coverage stays on the exchange, and he re-runs the numbers each fall at open enrollment, because earned income moves them ([[us-healthcare-and-social-security]]). - Four Social Security credits land, and each billing year replaces a zero in his 35-year average, so the benefit waiting at 67 inches up. He reads the new figure off his statement at the annual review. - The network stays alive, with two recurring clients. He keeps the option of going to $30,000 a year if a bad decade demanded it, sequence insurance whose strike price he knows. The lived cost is six weeks a year of work he enjoys at that dose. He tested it at double the dose during his prototype year, to find his limit. In the simulator, `Side income /yr` of $10,000 with `Side income until year` set to 15 brings the central failure probability down to 2.1%. The strongest plan in this book is not the biggest. It is the best matched. ::: ## The essentials - Work you choose is the best defensive asset in this book. The hour you work inside the fragile window is the best-paid hour of your financial life (the sequence effect). And the pay does not stop at the cash: health coverage, Social Security credits and a softened sequence ride along with it. - The forms are a choice. Barista (powerful, demanding on the work itself), Coast (gentle, slow), organic side income (the majority case, welcomed, never counted), the targeted return (pure anti-sequence). And the forced return, rare, which looks like a temporary Barista rather than like starting over. - The option **melts**, through skills, network and age. It is a first-decade margin. It is maintained (two to four weeks a year, a visible practice, a fed network, the paperwork ready) or it lapses, but knowingly, never by default. - The toolbox has five pieces: negotiated part-time where you already work (the one that carries benefits), W-2 part-time elsewhere, 1099 contracting (the highest rate, minus your own coverage and your own payroll tax), seasonal work, and volunteering. Open the entity **before** you leave. - Three stances hold together. Full FIRE (prudent rate, full margins), structural semi-FIRE (income sized and simulated), the maintained option (the recommended default, nothing counted, everything cultivated). And one test, where rate, floor and employability have to tell the same story. --- ## Going further - The Barista and Coast accounts in the corpus ([[voices-from-real-retirees]]) and the dedicated forum threads: the field evidence on the doses people actually run. - [ssa.gov](https://www.ssa.gov) for your earnings record and your benefit estimate. The credits, the 35-year average and the zeros are all on the statement, and it is worth a look before you set a date. - [healthcare.gov](https://www.healthcare.gov) for what a year of earned income does to the coverage side ([[us-healthcare-and-social-security]]). - Early Retirement Now on phased retirement and the "redo" ([[the-ern-series]]). - The academic file: Zhan, Wang, Liu & Shultz (2009) on bridge employment and health, Maestas, "Back to Work" (2010) on unretirement, Neumark, Burn & Button (2019) on age discrimination. - In this book: [[sequence-of-returns]] (why the early years), [[flexibility-in-practice]] (form number one), [[meaning-and-identity]] (work as usefulness you chose).