$144 a month. That’s the average U.S. residential electricity bill in 2024, drawn on 865 kilowatt-hours of usage, according to the U.S. Energy Information Administration. Add internet, gas, and water, and a shared household routes a few hundred dollars every month through whoever’s name happens to be on each account.
So the instinct is understandable: if we’re all paying, we should each pay for exactly what we use. Someone proposes an app that tracks each person’s kilowatt-hours, or a shared note logging shower minutes and thermostat changes. It sounds like the fairest possible system.
It’s usually the wrong one — not because precision is bad, but because most household utilities resist fair metering in ways that aren’t obvious until you’ve tried, and because the act of measuring a roommate quietly turns a home into a workplace. This guide is the case for splitting most utilities equally: the economics behind it, and the short list of costs that genuinely are worth itemizing.
Source: Residential electric bills in Hawaii and Connecticut are twice those in New Mexico and Utah, U.S. Energy Information Administration, 2025
How should roommates split utility bills?
Three methods, and every guide lists them. Split equally — the bill divided by the number of roommates. Split by usage — metering or estimating each person’s share. Split by a fixed estimate — a negotiated percentage, where the roommate who travels half the month pays 25% instead of 50%.
The standard advice treats usage-based splitting as the fairest of the three and equal splitting as the lazy default. For most households, that ranking is backwards. Usage-based looks fair on paper and corrodes in practice, because the measurement is either impossible — you cannot meter one person’s share of a shared router — or so invasive that collecting it costs the household more than it saves. The table below is where each method actually holds up.
Does metering actually save money?
It’s worth being honest about what metering does well, because it does something real: when people see their own usage on a bill, they use less.
The largest evidence on this comes from an EPA-sponsored study of water billing in multi-family housing, conducted by Aquacraft and released in 2004. The study’s framing of the problem is the whole point: in buildings where water is bundled into rent, there is “little or no incentive for the end user, the resident, to save water because there is no direct pricing signal.” Individually submetered units — billed for their own measured consumption — broke that pattern, and the study found submetering achieved statistically significant water savings of 15.3 percent. The detail that matters most: buildings that split the master bill by a formula — a ratio based on unit size or occupancy, never on actual use — were not found to yield water savings, because, in the study’s words, the allocation method “does not appear to provide an incentive… since the charge is based upon a predetermined formula and not on actual use.” Measurement changed behavior. Allocation formulas didn’t.
So if your only goal is to cut consumption, metering works. The real question is whether the saving is worth what it costs you to collect it — and between roommates, it almost never is. To see why, look at who was doing the monitoring.
Source: National Multiple Family Submetering and Allocation Billing Program Study, Aquacraft, Inc. for the U.S. Environmental Protection Agency, 2004
So why does metering backfire between roommates?
Here is the part the submetering studies don’t mention, because they weren’t measuring it: in every one of those buildings, a landlord billed a tenant. That is an arm’s-length, commercial relationship. Monitoring is expected; no one’s feelings are involved.
A roommate is not a landlord — and that difference is exactly what the economist Bruno Frey identified in his 1993 paper Does Monitoring Increase Work Effort? Frey showed that monitoring produces two opposite effects. In abstract, market relationships it has a “disciplining effect”: people watched more closely do more. But in personal relationships, monitoring has the reverse, “crowding-out” effect — being measured is perceived “as an indication of distrust,” and it erodes the intrinsic willingness to cooperate that held the arrangement together. His conclusion was specific about when each one wins: the crowding-out effect dominates when the relationship is personal; the disciplining effect dominates when it is abstract, as in a competitive market.
The hinge: The submetering studies caught the disciplining effect because they measured the abstract landlord–tenant relationship. Install the same monitoring between two friends and Frey’s research predicts it flips. The roommate who proposes tracking everyone’s kilowatt-hours isn’t buying lower usage — they’re buying a household where everyone now suspects everyone else of running the heat too high.
Frey and Reto Jegen later surveyed the wider evidence for this “motivation crowding” effect across dozens of studies in their 2001 review, and found it holds well beyond the workplace: external controls and incentives routinely undermine the internal motivation they were meant to reinforce. A shared home runs almost entirely on that internal motivation — the unmonitored, uncounted willingness to flip off a light, take a shorter shower, not run the dryer half-empty. Meter it, and you spend it. That’s the trade the savings never show: small gains in dollars, paid for in the one currency a household can’t refill.
Sources: Bruno S. Frey, “Does Monitoring Increase Work Effort? The Rivalry with Trust and Loyalty,” Economic Inquiry, 1993; Bruno S. Frey and Reto Jegen, “Motivation Crowding Theory,” Journal of Economic Surveys, 2001
Internet: the bill you should never meter
Some utilities can’t be metered fairly even in principle. Home internet is the clearest case — and the reason traces back to a definition from 1954.
That year, Paul Samuelson described what he called a “collective consumption good”: a good where one person’s consumption “leads to no subtraction from any other individual’s consumption” of it. Economists now call this property non-rivalry, and it’s the defining feature of a public good. A home internet connection is very nearly one. When your roommate streams a movie, the bandwidth available for your video call is unchanged — right up until the connection actually saturates, which for most households is rare. Your use and theirs aren’t in competition.
Why per-gigabyte splitting is a category error: There is no “fair share” of a non-rival good to measure, because no one’s use comes at anyone else’s expense. The connection costs the same fixed amount whether one person uses it or four. Splitting that fixed cost equally isn’t the lazy answer — it’s the economically correct one.
The same logic covers any shared subscription the household runs on top of the connection — the streaming bundle, the cloud storage, the music plan. Fixed cost, non-rival use, no meaningful way to measure individual shares. Split equally and move on.
Source: Paul A. Samuelson, “The Pure Theory of Public Expenditure,” The Review of Economics and Statistics, 1954
Electricity: the one utility worth occasionally adjusting
Electricity is the exception that tests the rule. Unlike internet, it’s genuinely rivalrous — the power your roommate’s space heater draws is power that shows up on a bill you share — and it’s the one utility that arrives already metered at the unit. So this is where usage actually could be tracked.
It still mostly shouldn’t be. The base load of a shared apartment — refrigerator, router, water heater, shared lights, the AC cooling rooms you both walk through — is genuinely joint, and per-person attribution of it is guesswork dressed up as precision. What’s worth adjusting for isn’t the toaster. It’s the rare load that is large, individual, and obvious. Each of these clears that bar:
A car charged at home is a steady, sizable draw attributable to one person that nobody else shares. The single clearest case for an itemized line.
One roommate running heat, AC, and equipment full-time while the other is at an office is a real, durable asymmetry. A flat monthly adjustment is fairer than 50/50 — and far easier than metering it.
A personal climate appliance running in one bedroom is a large, traceable draw the rest of the apartment never benefits from. Worth a side agreement.
Continuous high-draw equipment for one person’s hobby or side income isn’t a shared cost. Itemize it, or put that one device on its own plug-in meter.
Notice what these have in common: each is big enough to matter, clearly belongs to one person, and is visible without anyone playing detective. That’s the bar. Everything below it — who showered longer, who left a light on — costs more in goodwill to track than it returns in dollars.
Gas and water: base load with seasonal swings
Gas and water behave like electricity’s base load with none of its itemizable spikes. Most of a water bill is fixed service charges plus genuinely shared use — dishes, cleaning, the bathroom everyone uses. Gas swings hard by season: a low heating bill in autumn climbs steeply by deep winter. Neither rewards per-person tracking.
The one real trap is the roommate who joins or leaves mid-season and a bill that doesn’t reflect it — the person who moves out in February feeling they subsidized a winter they’re leaving. The fix isn’t metering, it’s averaging: split the seasonal total evenly across the months you actually overlapped, or true up once at lease-end. Don’t relitigate a steep January bill in the moment. Smooth it, and the seasonal swing stops being anyone’s grievance.
The fair rule: equal by default, itemize only the big three
Strip away the utility-by-utility detail and a single rule covers almost every shared bill: split it equally, and itemize a cost only when it passes all three tests at once. Here is the whole framework.
Default to equal
Most utilities are shared, non-rival, or too small to attribute, so equal split is the honest baseline — not the lazy one. Start here and make a cost earn its way off the default.
Is it large?
A line worth a side agreement is big enough that the dollars clearly cover the friction of tracking them. Below that, small asymmetries even out; chasing them doesn’t.
Is it clearly one person’s?
An EV, a home office, a personal appliance — an unambiguous owner. If you have to argue about whose it is, it’s effectively shared. Split it and stop.
Is it visible without surveillance?
If confirming the cost requires monitoring your roommate’s daily behavior, the monitoring is the cost — the one Frey’s research warns about. A charge that only a spreadsheet of shower minutes could prove never clears this test.
Large, attributable, visible — all three, or it’s a shared cost and it splits evenly. That one rule replaces every utility-by-utility debate, and it’s short enough to settle in the move-in conversation instead of the third month’s argument.
Who actually pays the bill?
Splitting fairly is only half the problem. The other half is mechanical: whose name is on each account, who fronts the money, and who chases the reimbursement. Left unassigned, all three quietly fall on one person — the household’s unpaid bookkeeper, usually the most organized roommate, who ends up carrying everyone’s float and doing the monthly math alone.
Three decisions close that gap. Put one name on each account and spread the accounts across roommates, so no single person carries all of them. Turn on autopay, so no bill depends on someone remembering. And pick one settle-up day a month, so balances never age past the point where “I’ll get you back” stops meaning anything. The full version of that setup conversation — rent, groceries, supplies, and all — is its own guide: how to set up expense splitting with new roommates. And when the split itself is what’s generating friction, the psychology of roommate money conflict covers how to raise it without a fight. The point here is narrower: the fairest split still fails if the admin lands on one person.
What splitty does here (and what it doesn’t)
A monthly utility ledger and a one-off shared cost are different problems, and they need different tools. Being honest about which is which is the whole point of fair splitting.
An expense tracker
Rent and the monthly utility reimbursement are a running ledger — the same accounts, every month, settled net. That’s what an app like Splitwise is built for, and it’s the right tool for the recurring side.
splitty
The Costco run you fronted for the apartment. The streaming bundle you paid for the house. The dinner the four of you went out for. Scan the receipt, assign what belongs to whom, and send each person a request before the moment passes — including that large, attributable load worth itemizing.
splitty’s honest take: we don’t track your monthly electric bill — and you shouldn’t want an app that surveils it. Split the recurring utilities equally and settle them on a tracker. Reach for splitty the day a real, shared cost lands and someone needs paying back. That’s the quick, one-off problem it solves.
FAQ
Frequently asked questions
01 How should roommates split utility bills?
Split most utilities equally. Internet, water, and the base load of an electric bill are shared, fixed, or too small to attribute per person, so an equal split is the fair baseline — not the lazy one. Reserve adjustments for a cost that is large, clearly one person's, and visible without monitoring, like an EV charging at home or a full-time work-from-home setup.
02 Should roommates track electricity usage per person?
Usually no. Most of a shared apartment's electricity is joint base load — refrigerator, router, water heater, shared AC — that can't be fairly attributed to one person. Per-person tracking also reads as distrust between roommates, which research on motivation crowding shows erodes cooperation. Adjust only for a large, individual, obvious load (an EV, a space heater, a home office), and use a flat monthly adjustment rather than metering.
03 How do you split an internet bill with roommates?
Equally. Home internet is a non-rival good: one roommate streaming doesn't reduce the bandwidth available to another until the connection actually saturates, which is rare. The connection costs the same fixed amount whether one person uses it or four, so there's no individual share to meter. The same applies to shared streaming and subscription bundles — fixed cost, split evenly.
04 What if one roommate uses way more utilities than the others?
Adjust only if the extra use is large, clearly attributable to that person, and visible without surveillance — an EV, a space heater, a home server, or a genuine full-time work-from-home asymmetry. In those cases a flat monthly adjustment is fairer and far less corrosive than metering. If the difference is small or hard to pin down, split equally and let it even out.
05 What's the best app for splitting bills with roommates?
Use two tools for two jobs. For recurring monthly costs — rent, the utility reimbursement — an expense tracker like Splitwise keeps a running ledger you settle net. For a one-off shared cost, like a Costco run you fronted or a dinner out, splitty scans the receipt, assigns each person's share, and sends pre-filled payment requests. Your roommates don't need to download anything.
