Most retirement income plans are built on a simple assumption: spending stays roughly flat, or rises steadily with inflation, from the day you retire until the day you don't. New research from J.P. Morgan Asset Management suggests that assumption is wrong, and the way it's wrong has real consequences for how you should structure your income in retirement.
J.P. Morgan's 2025 Retirement by the Numbers report, drawn from more than 12 million defined contribution participants and spending patterns from over 4.7 million retiree households, found that average retiree spending declines by more than 30% between ages 60 and 85. It also found that spending is far more volatile year to year, especially early in retirement, than most retirement models account for. Neither finding is what a typical retirement calculator assumes, and both point to the same conclusion: a retirement income plan built on a flat spending line is a plan built on the wrong shape.
Spending Declines, But Not on a Straight Line
The research found that retiree households spend an average of $75,630 per year between ages 60 and 64. By ages 90 to 94, that figure drops to $51,920. Spending trends down by roughly 5% to 8% every five years through most of retirement, then levels off in the late 80s and 90s as discretionary spending, travel in particular, naturally tapers.
This pattern makes intuitive sense. The early years of retirement tend to be the most active: travel, hobbies, helping adult children, home renovations. As retirees move into their 70s and 80s, discretionary spending typically declines, even as healthcare costs rise. The net effect, according to the data, is still a gradual decline in total spending, not the steady increase many retirement projections assume.
The planning implication is direct. If your retirement income model assumes spending grows with inflation every year for 30 years, it likely overstates what you'll need in your 80s and understates the importance of getting the first decade right. We cover how to build a withdrawal strategy flexible enough to absorb this kind of front-loaded spending pattern in Retirement Planning: What It Really Takes to Get It Right.
The Bigger Surprise: Year-to-Year Volatility
Even more significant than the long-term decline is how much spending fluctuates in any given year. According to the research, 60% of new retirees see their annual spending swing by 20% or more in the first three years of retirement, compared to their pre-retirement baseline. That volatility doesn't fully settle down with age, either: among retirees aged 75 to 80, 54% still experience swings of similar magnitude.
Michael Conrath, J.P. Morgan's chief retirement strategist, pointed to housing and healthcare costs as common drivers of these swings. A new roof, a move closer to family, an unexpected medical expense: any one of these can push a single year's spending well above or below the long-term trend, even while the multi-decade trajectory is downward.
This matters because sequence-of-returns risk, the danger of needing to withdraw a large sum during a market downturn early in retirement, is compounded when spending itself is unpredictable. A retirement plan that assumes smooth, predictable withdrawals doesn't hold up well against a reality where a fifth of your annual spending can shift in either direction with little warning.
Income Replacement Isn't the Same for Everyone
The research also challenges the common shorthand that retirees need to replace 70% to 80% of pre-retirement income. In practice, the replacement rate varies enormously by income level. Households earning $40,000 before retirement replace roughly 95% of that income in retirement, largely because Social Security covers a large share of the need. Households earning $300,000 replace closer to 55%.
What stays consistent across income levels is the share that has to come from savings, generally between 39% and 44%, regardless of whether a household earned $50,000 or $500,000 before retiring. Higher earners simply rely far less on Social Security as a percentage of their total income, which puts more weight on how well their investment and withdrawal strategy is designed.
There's a further wrinkle for anyone estimating Social Security with a generic online calculator: the research found that actual Social Security payouts run 20% to 35% lower, on average, than typical industry models suggest, with the gap peaking for households earning between $70,000 and $100,000 before retirement. That's a meaningful enough gap to change a retirement date. For pre-retirees weighing when to claim, and whether to draw down savings first to delay and maximize the benefit, we walk through the tradeoffs in The Social Security Bridge Strategy: How to Maximize Lifetime Income by Delaying Benefits. Gap-year income planning also intersects closely with Roth conversion timing, which we cover in Is a Roth Conversion Right for You?.
Debt Entering Retirement Is a Bigger Drag Than Many Assume
One of the more sobering findings involves debt. Nearly half of defined contribution plan participants, 48%, carry credit card debt, and the research found that higher balances are associated with lower contribution rates, smaller account balances, and a higher likelihood of borrowing against a retirement plan. For older participants, carrying credit card debt was associated with reducing retirement readiness by as much as 40%.
This is a reminder that retirement income planning doesn't start on the day you retire. It starts with the balance sheet you carry into it. Coordinating debt paydown with retirement contributions, rather than treating them as separate problems, is one of the areas where a tax planning strategy and a retirement plan need to work together. We cover this coordination in more depth in Tax Planning Strategies That Actually Move the Needle.
What This Means for Your Retirement Income Strategy
Taken together, these findings argue for a different kind of retirement income plan than the flat, inflation-adjusted withdrawal model many people picture.
- Build flexibility into your withdrawal strategy. If spending genuinely swings 20% or more in a given year, your plan needs a way to absorb that swing without forcing a bad decision, such as selling equities during a downturn to cover an unplanned expense.
- Weight the early retirement years carefully. Since spending tends to be highest in the first decade and sequence-of-returns risk is most damaging during that same window, asset allocation in the years just before and after retirement deserves particular attention.
- Model your actual Social Security benefit, not a rule-of-thumb estimate. Given how far actual payouts can diverge from generic calculators, this single number can shift a claiming decision by years.
- Address debt as part of the retirement plan, not separately from it. Entering retirement with high-interest debt undermines the income strategy built around it.
- Revisit income replacement assumptions based on your actual income level, not a generic percentage. Higher earners in particular should expect to fund a larger share of retirement from savings rather than Social Security.
None of this is a reason to abandon planning. It's a reason to plan with a model that reflects how retirement spending actually behaves rather than how it's assumed to behave. This is the kind of coordinated, ongoing planning we build into every retirement engagement at MJT & Associates. For a broader look at how retirement income, tax strategy, and legacy planning fit together, see A Real Financial Plan: What It Includes, Why It Matters, and Where Most People Fall Short.
Frequently Asked Questions
Q1: If retiree spending declines over time, does that mean I need less saved than I thought?
Not necessarily. The decline in average spending doesn't eliminate the need for a healthy cushion. Healthcare costs tend to rise even as discretionary spending falls, and the year-to-year volatility the research identified means you need enough flexibility to absorb an unusually high-spending year at any point in retirement, not just a lower average over three decades.
Q2: I'm five years from retirement. What should I actually do with this information?
Focus on the years immediately surrounding your retirement date. Since spending tends to be highest early in retirement and market losses do the most damage when withdrawals are being taken at the same time, this window deserves a close look at your asset allocation, your cash reserve, and how much flexibility your income sources give you if a given year runs well above plan.
Q3: How much should I trust an online Social Security estimate?
Treat it as a starting point, not a number to build a claiming decision around. The research found real-world payouts frequently run meaningfully lower than generic calculator estimates, particularly for middle-income earners. A precise projection based on your actual earnings record is worth the time before you commit to a claiming age.
Q4: Does this research change how much I should be contributing now if I'm not close to retirement?
It reinforces the value of starting early. The same research found that increasing contributions by just 1% starting at age 25 can fund roughly nine years of average Medicare-related expenses in retirement. Small, early increases compound into meaningful retirement income capacity.
Related Reading on the MJT Blog
The Social Security Bridge Strategy: How to Maximize Lifetime Income by Delaying Benefits
Retirement Planning: What It Really Takes to Get It Right
Is a Roth Conversion Right for You?
A Real Financial Plan: What It Includes, Why It Matters, and Where Most People Fall Short
Tax Planning Strategies That Actually Move the Needle
Conclusion
Retirement spending doesn't move in a straight line, and neither should your income plan. The research is clear that spending declines over the long run, swings considerably from year to year, and depends heavily on your specific income level and Social Security record, not a generic rule of thumb. A plan built to absorb that reality, rather than one built on a flat assumption, is what actually holds up over three decades of retirement.
At MJT & Associates, we build retirement income plans around how spending actually behaves, not how it's assumed to behave on a spreadsheet. That means stress-testing your plan against real volatility, modeling your actual Social Security benefit, and revisiting the plan as your spending evolves.
Ready to see how your retirement income plan holds up against real-world spending patterns? Contact us today to schedule a consultation.
