The federal energy credit expired. Only three projects on your list should change.

It is late September and you are standing in the basement of a house built in 1954, holding two quotes you collected in the spring. One is for a heat pump. One is for air sealing and attic insulation. Both of them have a line near the bottom that says something like less federal tax credit — and both of those lines are now wrong.

You did not do anything wrong. The quotes were accurate when you got them. What changed is that the federal subsidy underneath them expired, the contractors’ proposal templates did not, and nobody sent you a letter about it. So the first cold night of the season arrives and the question you actually have is not “should I do this project” but “does my whole plan still work.”

Mostly, yes. But you should know exactly which parts moved.

What actually changed

Two federal credits ended on the same day.

The Energy Efficient Home Improvement Credit — 25C, the one that covered insulation, air sealing, exterior doors, windows, heat pumps, and heat pump water heaters — is, per the IRS, “allowed for qualifying property placed in service on or after Jan. 1, 2023, and before December 31, 2025.” It was worth up to $3,200 a year at 30 percent of cost, split into a $1,200 general bucket (with sub-caps of $600 for windows, $500 for doors, $150 for a home energy audit) and a separate $2,000 bucket for heat pumps, heat pump water heaters, and biomass systems.

The Residential Clean Energy Credit — 25D, the uncapped 30 percent credit on rooftop solar, batteries, and geothermal — ended on the same date. The Congressional Research Service is blunt about the mechanics: an expenditure counts when the original installation is completed, so a system finished after December 31, 2025 gets nothing, regardless of when you signed or paid.

There is no phase-down and no transition window. This is the first full fall — the first real heating-season planning cycle — on the other side of it.

Why this lands hardest on old houses

Because old houses are where the credit-eligible project list lives.

The median owner-occupied home in America is now 42 years old, up from 31 in 2005. Roughly 47 percent were built before 1980 and 34 percent before 1970 — which means the median American homeowner owns a house built before the modern energy code, and a third of us own houses built before anyone was really thinking about an envelope at all.

Those houses cost more to keep. Harvard’s Joint Center for Housing Studies finds owners of pre-1960 homes spent an average of $6,000 on improvements and maintenance in 2023, against $4,500 for homes built in 2010 or later — and that the gap between income quintiles is enormous, $12,700 at the top against $3,400 at the bottom. The Federal Reserve Bank of Philadelphia’s 2025 home repair cost estimates put it more starkly: 48 percent of owner-occupied homes built before 1940 needed at least one repair in 2024, against 26 percent of homes built in 2000 or later.

So the cohort with the longest project list, the thinnest margin, and the least reversible decisions is the cohort that just lost the subsidy. That is worth naming plainly. It is not, however, a reason to tear up the list.

The credit was ranking your projects, and that was always the problem

Here is the part nobody says out loud. For four years, the federal credit was not just paying for work on old houses. It was sorting the work.

Think about how the advice actually reached you. A contractor’s proposal put the credit-eligible line items at the top, because that is where the discount showed. A utility’s website led with heat pumps. Every “what should I do to my old house” article you read between 2023 and 2025 opened with insulation and a heat pump, because that is what the money was pointed at. None of that was dishonest. It was just a ranking produced by federal policy goals — decarbonization, load electrification, grid planning — rather than by the specific failure risks sitting in your specific basement.

Most of the time those two rankings agreed, which is why it worked so well and why so few people noticed the substitution. Air sealing an under-insulated 1954 house is genuinely the best first dollar you can spend on it, credit or no credit. But sometimes they disagreed, and when they disagreed, the credit won — because a 30 percent discount is a very loud argument and “the cast iron drain stack has nine years left” is a very quiet one.

Now the loud argument is gone. What is left is your house. That is not a downgrade.

Three projects that genuinely change

Be honest about where the economics actually moved. Three items, and only three.

1. Whole-house window replacement, done for energy reasons. Windows were always the weakest energy project on the list — the highest cost per unit of comfort gained, and the one the credit flattered most. Note that 25C capped windows at $600 a year against a project that routinely runs $15,000 to $30,000 on an older home; that is a 2-to-4 percent discount dressed up as a tax incentive. Losing it barely changes the arithmetic, and that is the point: the arithmetic was never good. If you were replacing windows to save energy, stop, and put the money into air sealing and attic insulation, which will outperform it by a wide margin. If you are replacing windows because the sashes are rotted, the glazing is failing, the sills are letting water into the framing, or a second-floor bedroom has no code-compliant egress — that is not an energy project. It never was. Do it on the failure-risk clock, not the subsidy clock.

2. Rooftop solar. This is the real change. Twenty-five D was 30 percent, uncapped, on a project that commonly costs $20,000 to $35,000 installed — the largest single subsidy most old-home owners had access to, and its removal materially lengthens payback. If a cash purchase penciled at a nine-year payback with the credit, run the new number before you assume it still works; it will not be nine years. Two things worth knowing before you conclude solar is dead. First, 25D was the homeowner’s credit; the separate business-side credit that underwrites third-party ownership — leases and power purchase agreements — is on its own schedule and has not ended, so the economics of not owning the array have changed much less than the economics of owning it. That schedule is also being narrowed, so ask any installer quoting you a lease to put in writing which credit their pricing assumes and through what date. Second, a lease is a twenty-year contract on your roof, which makes it downstream of a question it did not used to be downstream of: how old are your shingles? Never put an array on a roof with fewer years left than the contract.

3. Early replacement of a fossil system that still works. The $2,000 heat pump credit was, in practice, the thing that made “replace the working furnace now” beat “replace it when it dies.” That was the credit’s whole job on this item, and it did it well. Without it, for most old houses, the right answer reverts to: run it to failure, but pre-decide the replacement now. One exception, and it is not negotiable — if a technician finds any condition that can put combustion gas into your living space, the furnace is done that day. That is a conditional sentence with a cost attached: nobody finds those conditions unless somebody looks, so “run it to failure” is only honest if you are booking the annual combustion-safety inspection in the October list below. A cracked heat exchanger is the example everyone has heard of, and it is only one of several: a flue that is blocked, disconnected, or corroded through; a furnace that backdrafts under negative pressure, which is a live risk in a house that was recently air-sealed; a burner overfiring. If the technician red-tags the unit or tells you not to run it, that is the end of the conversation. Any of these is a carbon monoxide finding, not a maintenance finding, and “run it to failure” does not apply to a failure mode that can kill you in your sleep. Get the load calculation done, size the equipment, identify whether your panel and your ductwork can carry a heat pump, and get the quote — this fall, while nobody is in crisis. Then when the furnace dies in January you make a phone call instead of a decision. Pre-deciding costs a few hundred dollars and buys you the single most valuable thing in an emergency: the ability to not be talked into something in a cold house.

Everything else on your list did not move

Which is most of it. The credit never touched it, so its expiration cannot touch it either.

Air sealing and attic insulation. Still the best first dollar in an old house, by a distance. Blower-door-guided air sealing plus insulation to current attic depth is the one project that improves comfort, reduces energy cost, and makes the eventual HVAC equipment smaller and cheaper all at once. The credit covered a sliver of it. Its absence changes nothing about its rank.

Anything on a failure clock. Roof at year 22. Electrical panel that is a Federal Pacific Stab-Lok or a Zinsco. Knob-and-tube in a wall you are about to open. Water heater at year 13. Cast iron stack that is scaling shut. None of these were ever credit-eligible, all of them are on their own schedule, and none of them care what Congress did.

Anything on a water clock. Grading that falls toward the house, gutters that discharge at the foundation, a failed sump, a chimney with no cap. Water is the only failure mode that reliably turns a $2,000 problem into a $40,000 problem while you are not looking. This is still the cheapest insurance in an old house.

Anything coupled to access. The rule in an old house is that the expensive part is rarely the material — it is opening the wall. If you are going to open a wall for one reason, do every job inside that wall at once. Rewire, insulate, rough-in the plumbing, and add the blocking, all in the same opening. Sequencing around access will save you more money than any tax credit ever did.

The replacement money is a lottery. Check it last.

There is federal money still moving, but it does not behave like a tax credit and you should not plan around it like one.

The Home Energy Rebates — the $8.8 billion HOMES and HEAR programs — are administered by individual states, not by the IRS. That difference matters more than the dollar amounts. A tax credit is an entitlement: if you did the work and you owed the tax, you got it. A state rebate is a queue. Programs launch on each state’s own timeline, run against a fixed allocation, are usually capped by household income, and close when the allocation is gone. As of this month some states are live, several plan to open later this year, and some have no announced launch date at all. California’s single-family electrification reservations were fully subscribed and moved to a waitlist earlier in 2026.

Your utility is the other real source, and often the better one — rebates on heat pumps, water heaters, insulation, and free or subsidized energy audits, with no federal strings. DSIRE is the cleanest way to see what your state and utility actually offer at your address.

Use all of it. But use it in the right order. Decide the project from the house, then go looking for money. If you invert that — if you let the availability of a rebate choose which project you do — you have handed your capital plan to a queue, and a queue is a worse project manager than the tax credit was.

What to do in October

Six things, none of which take more than a weekend.

Test the carbon monoxide alarms, and make sure one of them is outside the bedrooms. This is the cheapest item on the page and the one everything above it rests on, because “run the furnace to failure” is only a sane plan in a house that will tell you when combustion gas shows up. Press the test button on every unit, then read the manufacture date on the back: the sensors expire, usually at seven to ten years, and an expired alarm fails silently. The CPSC rule is one on every level of the home and one outside the sleeping areas — which in a two-story 1954 house with the furnace in the basement means three alarms, not two. Most people install fewer. A replacement is about $25.

Book the annual combustion-safety inspection. This is the trigger the whole “run it to failure” plan depends on. CPSC urges “a yearly professional inspection of all fuel-burning home heating systems, including furnaces, boilers, fireplaces, wood stoves, water heaters, chimneys, flues and vents.” Ask for the four things that decide whether the furnace is safe to keep running: the heat exchanger, the integrity of the flue and its connectors, draft under the worst case your house can produce, and the burner’s firing rate. Say “including the venting” out loud when you book it: CPSC’s instruction is that the yearly inspection “should include chimneys, flues, and vents for leakage and blockages by creosote or debris,” and the venting is the part that fails quietly. Without this visit, nobody is ever going to find the condition in the exception above, and your alarms are your entire safety layer.

Re-price the two or three quotes you are sitting on. Any proposal written before this year that shows a federal credit line is overstating your savings by the amount of that line. Ask the contractor to reissue it without the credit. You want the real number on the page before you compare anything.

Get a blower-door energy audit and pay for it yourself if you must. The $150 credit for audits is gone, and many utilities still subsidize or cover them. This is the single highest-information purchase available on an old house — it tells you where the envelope is actually leaking rather than where you assume it is, and it turns “the upstairs is cold” into a ranked list with numbers on it.

Pre-decide the HVAC replacement you are not making yet. Load calc, equipment selection, panel and duct check, written quote. Put it in a folder. Do not install it.

Then, and only then, go look for money. Check DSIRE, check your utility, check whether your state’s HEAR or HOMES program is open and whether you are under the income cap. If money exists for something already on your list, take it. If it exists for something that is not on your list, that is not a reason to add it.

Why this is a decision-support problem

The reason we write about this is that the hard part of owning an old house has never been knowing what a house needs. Roofs age, panels fail, envelopes leak, and none of that is secret. The hard part is that six different parties are each pointing at a different item on your list, and every one of them is answering a question that is not quite yours.

For four years, one of those parties was the federal tax code, and it was the most persuasive one in the room. It has left. That feels like a loss, and on solar it genuinely is one. But on most of the list, what actually happened is that a loud, well-intentioned, somewhat arbitrary ranking stopped drowning out the quiet one your house has been offering the whole time.

Your sequence was never supposed to come from a tax form. It comes from what fails first, what that failure takes with it, what becomes unreachable once the wall closes, and what you can actually spend this year. Work that list. Then go see whose money you can put on top of it.