Projecting Your Retirement Income Needs

Projecting retirement income needs is easier when the question is framed as a spending problem rather than an income-replacement problem. A household earning $150,000 shortly before retirement does not automatically need a fixed percentage of that salary after work ends, because part of the salary may currently be going to retirement contributions, payroll taxes, commuting, debt payments or expenses that will disappear. At the same time, retirement can introduce costs that were smaller or absent during the working years, including Medicare premiums, additional travel, home assistance or higher out-of-pocket health spending.

Rules of thumb such as replacing 70% or 80% of pre-retirement income can be useful for a quick first estimate, but they should not be mistaken for a personal target. The better starting point is to work out how much money you need in retirement to support the life you actually expect to lead, then compare that spending target with Social Security, pensions, part-time earnings and the amount that must come from savings. The result is not one permanent number. It is a planning range that should be updated as retirement gets closer and actual expenses become clearer.

Start with spending, not a salary percentage

The old version of this article made a useful distinction between essential expenses and discretionary lifestyle spending. That idea should remain, but it works better as part of a complete retirement budget rather than as an exercise in deciding what is a true “need.” Housing, food, utilities, transportation, insurance and basic health care form the core of the budget, while travel, dining out, hobbies, gifts and other discretionary spending determine how much room the plan needs for lifestyle choices. Both categories matter because a retirement plan that covers only bare necessities may be technically sustainable but may not resemble the retirement the household actually wants.

Start with recent spending records rather than gross income. Bank and credit-card statements, insurance bills, tax returns and recurring-payment records usually provide a better picture than memory. Twelve months is a reasonable minimum because property taxes, insurance renewals, vacations, annual subscriptions and home maintenance do not arrive evenly each month. Two or three years of records are even more useful when spending has been unusually high or low in one year.

National spending data can provide context, but it cannot substitute for the household budget. The Bureau of Labor Statistics publishes Consumer Expenditure Survey tables by age, income, household size, housing tenure and other characteristics, which shows why a single replacement ratio cannot describe every retiree’s spending pattern.^[1] A mortgage-free couple in a lower-cost area can have a very different retirement budget from a renter in an expensive city even when their pre-retirement salaries were similar.

When reviewing current spending, separate expenses that are likely to continue from those tied directly to work. Commuting, professional clothing, lunches bought near the office and some payroll-related costs may fall, while employer-paid benefits can disappear. Retirement contributions also stop being an expense once saving turns into spending. A household that currently directs 15% of income into a 401(k) does not need to replace that contribution as retirement consumption, although it still needs enough assets to produce the income that saving was intended to fund.

Debt needs its own line because the timing matters. A mortgage scheduled to end three years after retirement creates a higher income requirement at first and a lower one later. Car loans may disappear and then return when a vehicle is replaced. Credit-card debt and other high-cost borrowing should not simply be assumed to continue indefinitely, because carrying it into retirement can make the required income much larger and reduce the amount of the portfolio available to support spending.

Build the retirement budget in today’s dollars

A practical first projection is usually easiest to build in today’s dollars. Estimate what the desired retirement lifestyle would cost at current prices, then let the retirement calculation apply inflation consistently. Mixing today’s rent, a future Social Security estimate and a healthcare number inflated for ten years can produce a projection that looks precise but is internally inconsistent. The plan should either keep all amounts in real, inflation-adjusted terms or inflate all future cash flows to the year in which they are expected.

Housing is often the largest category and deserves more detail than “mortgage or rent.” Homeowners still face property taxes, insurance, utilities, repairs and major replacements after the mortgage is gone. A roof, heating system, appliance or accessibility renovation can create a large one-time expense that an ordinary monthly budget misses. Renters avoid many repair costs but remain exposed to rent increases and may want a larger housing buffer if they expect to stay in a high-cost market.

Transportation costs often fall after commuting ends, but they do not disappear. Insurance, fuel, maintenance, registration and eventual vehicle replacement remain part of the long-term budget. A household that expects to reduce from two vehicles to one should model the saving explicitly rather than assuming transportation will simply become “lower.” If retirement includes more road trips or relocation to an area where driving is essential, the cost can move in the opposite direction.

Food and ordinary household spending are relatively easy to estimate from current records, but discretionary categories deserve realistic treatment. Travel, restaurants, hobbies and entertainment are not automatically wasteful merely because they are optional. They are often part of the reason people save for retirement in the first place. The useful question is how much of that spending is affordable under the base plan and how much could be reduced temporarily if markets, health or other circumstances make a lower spending level necessary.

Family support should also be budgeted when it is likely rather than treated as an unexpected exception every year. Gifts to children or grandchildren, tuition assistance, help with a home purchase and recurring support to relatives can become meaningful retirement outflows. If those commitments are important, putting them into the projection makes the trade-off visible. If they are aspirational, they can be treated as a secondary goal that is funded only when the primary retirement plan remains on track.

Large irregular expenses are best handled separately from the normal monthly budget. Home repairs, vehicle replacement, dental work, hearing aids, major travel and other episodic costs can be represented as scheduled or contingency amounts rather than smoothed into a monthly figure that hides their timing. A retirement plan is more useful when it can show that a $30,000 expense in year six is different from adding $2,500 of recurring annual spending forever.

Estimate how expenses will change after work ends

Some retirement costs decline quickly, some rise, and others change only later. A projection that assumes one constant percentage of salary often misses these transitions. The first several years of retirement may include travel and other active-lifestyle spending, while later years may involve less travel but more spending on health care, assistance or changes to the home. There is no universal spending curve, so the plan should reflect the household’s own expectations rather than assume that retirement becomes cheaper every year.

Health care deserves a separate estimate because employer coverage, Medicare and out-of-pocket costs work differently. Someone retiring before Medicare eligibility may need to budget for marketplace coverage, COBRA or a spouse’s employer plan, while Medicare-eligible retirees still face premiums, deductibles, copayments, prescription costs and services that Medicare does not fully cover. The cost of health care insurance is therefore not one number that can simply be copied from the final working-year paycheck.

For 2026, the standard Medicare Part B premium is $202.90 per month, and higher-income beneficiaries can pay more. Medicare Advantage, Medigap and Part D costs vary by plan and location, while Original Medicare also leaves deductibles and coinsurance to be considered.^[2] A retiree who is still several years away from Medicare should not build the plan around the current dollar amount, but the current structure is a reminder that “Medicare” does not mean health spending falls to zero at age 65.

Taxes change as well. Wages may be replaced by Social Security, pension income, traditional retirement-account withdrawals, Roth withdrawals, interest, dividends and realized capital gains, and those sources do not all have the same federal tax treatment. A retirement budget stated as after-tax spending therefore needs to be converted into the gross income required to support it. The household should avoid comparing an after-tax spending target with gross Social Security and pension amounts and then treating the difference as the portfolio withdrawal need.

State taxes and location can materially change the projection. A move to another state may affect income taxes, property taxes, housing costs, insurance and health-plan choices at the same time. Moving to a lower-cost area can reduce the required retirement income, but the saving should be based on a real destination budget rather than a broad assumption that another state or country will be “cheaper.” Travel back to family, housing transactions and different insurance costs can absorb part of the apparent advantage.

Map reliable income before calculating the portfolio gap

Once the spending target has been estimated, the next step is to identify the income that does not depend on selling investments. Social Security, defined-benefit pensions, annuity payments and part-time work may cover part of the budget. The timing matters because these income sources may begin in different years, and a household can face a temporary gap between retiring and starting Social Security or a pension.

Social Security should be entered using an individualized estimate rather than a national average. A my Social Security account can show projected retirement benefits at different claiming ages based on the worker’s earnings record, and the estimate should be updated as earnings and claiming plans change. Couples should examine both records, survivor implications and the possibility that benefits begin at different times instead of simply adding two current estimates together.

Pensions need similar care. The monthly benefit may depend on retirement date, years of service, final compensation and the survivor option selected, and some pensions do not receive full inflation adjustments after payments begin. A joint-and-survivor option generally produces a different monthly benefit from a single-life option because it can continue income for a surviving spouse. The retirement-income projection should use the election that is realistically expected, not the largest number shown on a benefit statement.

Part-time work can reduce the amount that must be taken from savings, but it should not be treated as guaranteed income unless the work is genuinely dependable. A plan that requires substantial earned income into the late seventies is different from one in which work is optional. Health, caregiving responsibilities and the labor market can all affect the ability to keep earning, so employment in retirement is often better modeled as a helpful scenario rather than the only way the plan succeeds.

The difference between required spending and reliable after-tax income is the amount the portfolio must support. This is the point at which retirement savings move from an accumulation problem to an income problem. If the gap is $40,000 in the first year, that does not mean the portfolio must generate exactly $40,000 of interest or dividends. Withdrawals can come from a combination of income, matured fixed-income holdings and asset sales, provided the overall withdrawal strategy remains compatible with the portfolio and the length of retirement.

Allow for inflation and a long retirement horizon

A retirement-income target is incomplete without a time horizon. A household retiring at 55 asks its portfolio to do a different job from one retiring at 70, even if first-year spending is identical. The earlier retiree faces more years of withdrawals, more years in which inflation can compound and a longer period in which poor investment returns can affect the plan. Delaying retirement by even a few years can improve the calculation in several ways because savings have more time to grow and the period funded by withdrawals becomes shorter.

Average life expectancy is useful context but should not be used as an expiration date for the plan. In the Social Security Administration’s 2023 period life table used for the 2026 Trustees Report, a 65-year-old man has an average remaining life expectancy of 18.12 years and a 65-year-old woman 20.66 years.^[3] Those are averages, not planning guarantees. A substantial share of retirees will live longer, and couples also face the possibility that one spouse survives well beyond the other’s life expectancy.

Inflation creates a separate risk because a fixed nominal income buys less over time. If the budget is expressed in today’s dollars, the retirement model should use real returns or otherwise increase future spending assumptions with inflation. Social Security includes cost-of-living adjustments, but many private pensions and annuities do not provide the same protection. A household with a large fixed pension may therefore appear very well funded in the first decade and gradually rely more heavily on the portfolio as prices rise.

Inflation does not affect every category equally. Housing costs can be relatively stable for a mortgage-free homeowner until taxes, insurance or repairs rise, while rent can reset more quickly. Health care, travel, food and energy follow their own price patterns. A useful projection does not need a separate inflation forecast for every line item, but it should avoid pretending that a single assumption makes the future predictable.

Investment returns are equally uncertain. The expected return used in a calculator is an assumption, not an income promise, and the order in which returns occur can matter greatly once withdrawals begin. Poor markets early in retirement can force a larger percentage of the remaining portfolio to be sold for spending, even if long-run average returns later recover. This is one reason the investments supporting retirement income should be evaluated for liquidity, diversification and risk rather than selected solely for the highest expected return.

Test the plan with more than one scenario

A single retirement projection is most useful as a base case, not as a prediction. The household should know what happens if retirement begins earlier than planned, if spending is higher, if inflation stays elevated for several years, if markets perform poorly near the retirement date or if one spouse lives much longer than expected. The objective is not to create dozens of pessimistic scenarios. It is to identify which assumptions have enough influence to change the decision.

Start by testing the expenses that are genuinely flexible. Travel, gifts, dining and some large purchases can often be delayed or reduced without threatening basic living standards. Housing, health care, insurance and taxes are harder to cut quickly. Separating those categories gives the retirement plan a spending floor and a higher lifestyle target, which is more informative than treating every dollar as equally necessary.

That distinction also makes the old article’s essential-versus-discretionary idea more useful. The base plan should be able to support core expenses with a reasonable margin, while discretionary spending can expand when the portfolio and income sources allow it. The target level of comfortable retirement is then explicit rather than being hidden inside a percentage of salary. If the desired lifestyle requires aggressive return assumptions just to make the numbers work, the solution is usually to revisit retirement date, saving, spending or income rather than to treat higher investment risk as guaranteed compensation for a shortfall.

A household should also test the effect of one spouse dying earlier. Some expenses decline, but many do not fall by half, and income can change sharply because one Social Security benefit may end and pension payments may depend on the survivor election. Taxes can also change when the survivor eventually files as a single taxpayer. A plan that looks comfortable for two people should therefore be checked for the surviving spouse rather than assuming the household balance sheet remains unchanged.

Long-term care is another uncertainty that should be acknowledged without pretending to forecast an exact bill decades in advance. Some households choose insurance, others plan to self-fund, and many use a combination of assets and public programs depending on eligibility. The important point for an income-needs projection is to decide whether a separate reserve is needed and whether the ordinary annual retirement budget would still work if substantial care costs arose.

Turn the income target into a savings target

After the retirement budget and reliable income sources have been estimated, the remaining gap can be translated into a required portfolio. Retirement calculators do this by making assumptions about investment returns, inflation, taxes, retirement length and withdrawals. Different calculators can produce different answers because their assumptions differ, so the result should be examined rather than accepted simply because the tool displays a precise dollar amount.

For the accumulation side of the plan, the Retirement Savings Calculator can project a saving path or estimate the starting monthly saving associated with a target.

A useful calculation keeps the spending target and the portfolio gap separate. Suppose a household estimates that it needs $86,000 a year after tax in today’s dollars and expects Social Security and a pension to provide $48,000 after estimated taxes once both have begun. The long-term portfolio gap is about $38,000 a year, but the first several years could require more if retirement begins before those benefits start. The required savings level must therefore support both the temporary bridge period and the later recurring gap.

The example also shows why multiplying annual spending by a fixed number can mislead. Two households with the same $86,000 spending target can require very different portfolios if one receives a $50,000 inflation-adjusted pension and the other has no pension. Likewise, a household retiring ten years earlier may need a larger portfolio even if annual spending is lower. Retirement income needs and retirement savings needs are related, but they are not interchangeable.

Taxes should be included before the savings target is finalized. Withdrawals from traditional retirement accounts are generally taxable, qualified Roth withdrawals receive different treatment, and taxable investment accounts can produce interest, dividends and capital gains. If most savings are in pretax accounts, the portfolio may need to distribute more than the after-tax spending gap. The withdrawal plan and the income-needs projection should therefore be coordinated with retirement tax planning rather than developed separately.

Return assumptions deserve the same discipline. An expected return is useful for modeling, but increasing it solely to make a savings shortfall disappear does not improve the retirement plan. Higher-return assets normally bring more uncertainty, and the consequences of losses become more serious when the portfolio is simultaneously funding withdrawals. The savings target should be compatible with a portfolio the household can realistically hold through difficult markets.

Review the projection as retirement gets closer

A retirement-income estimate made twenty years before retirement cannot be expected to remain accurate without revision. Income changes, mortgages are refinanced or repaid, children become independent, health circumstances change and retirement preferences become more concrete. Early projections are still valuable because they reveal whether saving is broadly on track, but the plan should become more detailed as the retirement date approaches.

Five to ten years before retirement, the household should be able to replace broad salary percentages with an expense-based budget and more reliable Social Security and pension estimates. At that point, housing plans, debt payoff dates, health coverage before Medicare and likely retirement location should be sufficiently clear to model explicitly. Testing the budget for several months before retiring can expose categories that were underestimated or forgotten.

The final years before retirement also provide an opportunity to build liquidity for near-term spending. The portfolio does not need to produce every dollar of retirement income as yield, but money expected to be spent soon should not depend entirely on selling volatile assets at a favorable price. The exact amount of cash and short-term holdings depends on the withdrawal strategy, other reliable income and risk tolerance, so it should be determined as part of the broader portfolio plan rather than by a universal rule.

Once retirement begins, the projection becomes a living spending plan. Compare actual expenses with the estimate, update Social Security and pension income, account for taxes, and adjust discretionary spending when circumstances change. A plan that is reviewed can absorb surprises without treating every deviation as a failure, and it can also show when the household is spending less than it can reasonably afford.

The purpose of projecting retirement income is not to discover one perfect number decades in advance. It is to understand the level of spending the household wants to support, identify the income that is likely to be available and measure the gap that savings must fill over a long and uncertain period. When the assumptions are visible, retirement decisions become easier to test, and changes in spending, work, housing or retirement age can be evaluated in terms of what they actually do to the plan.