Roughly one in six U.S. households reported being unable to pay an energy bill in full over the past year, according to the Census Bureau’s Household Pulse Survey. That figure sits at the center of a long-running federal data trail showing that millions of Americans routinely choose between keeping the lights on and covering food, medicine, or rent. With the survey relaunched in January 2025 under the renamed Household Trends and Outlook Pulse Survey program, the data collection is active again, and the results point to a problem that has shifted in scale but never disappeared.
Energy prices, not jobs, appear to drive arrears
The one-in-six finding draws on the ENERGYBILL variable inside the Census Bureau’s HPS00 indicator group, which asks whether a household was unable to pay an energy bill in full during the preceding 12 months. That measure captures a narrower slice of hardship than earlier federal surveys. The Energy Information Administration’s 2015 Residential Energy Consumption Survey found that about one in three households faced challenges paying energy bills that year, a category broad enough to include receiving disconnection notices or making tradeoffs between utilities and other essentials. By 2020, the EIA reported that roughly 27 percent of households had difficulty meeting their energy needs, a decline from the 2015 benchmark but still a strikingly large share of the population.
The trajectory from one in three to roughly 27 percent to one in six might look like steady improvement, but the comparison is complicated by differences in survey wording and scope. The 2015 and 2020 EIA figures measure broad energy hardship, including difficulty and tradeoffs, while the Census ENERGYBILL variable isolates actual nonpayment. A household that skipped meals to keep the heat running would count in the EIA data but might not register in the Census measure if the bill was ultimately paid. The narrower Census question, in other words, likely captures the most severe cases of energy insecurity.
One hypothesis worth testing against the data is whether post-2022 energy-price spikes explain more of the variation in ENERGYBILL responses than shifts in unemployment or the end of pandemic stimulus payments. Natural gas and electricity prices rose sharply through 2022 and into 2023, and those increases hit household budgets directly. Unemployment, by contrast, remained historically low during much of that period. If state-level ENERGYBILL responses track more closely with local energy-price movements than with labor-market conditions, the policy implication is clear: job growth alone does not solve utility debt when the cost of heating and cooling a home keeps climbing.
That framing also helps explain why energy-bill distress can remain stubbornly high even when headline economic indicators look strong. A tight labor market may boost wages, but if regulated and unregulated utility rates rise faster than paychecks, the share of households falling behind can stay flat or even increase. The Census data, by focusing on missed payments rather than perceived difficulty, offers one way to distinguish between general financial strain and outright inability to keep up with essential services.
Federal data and credit records tell different parts of the story
The Census survey is self-reported, which means it reflects what households say about their own payment struggles rather than what utility companies record on their ledgers. No federal dataset currently merges the Census responses with utility-company arrears or disconnection records at the household level. That gap matters because self-reported hardship and actual account delinquency can diverge: some respondents may understate problems out of stigma, while others may report difficulty even when payments are technically current.
Credit-bureau data offers a partial cross-check. The Consumer Financial Protection Bureau found that items in collection on credit reports declined by about one-third from 2018 to 2022, a trend driven by changes in medical-debt reporting rules, pandemic-era relief, and stronger household balance sheets. Yet that improvement in recorded collections does not necessarily show up as a comparable drop in ENERGYBILL distress. Many utility accounts never appear on credit files unless they are sold to third-party collectors, and some utilities paused shutoffs or offered extended payment plans during the pandemic without immediately reporting delinquencies.
The result is two overlapping but distinct portraits of financial strain. Credit records capture debts that have escalated far enough to trigger collection activity, while the Census survey detects earlier-stage trouble: the moment a family realizes it cannot pay the current bill in full. In practice, a household might answer “yes” to the ENERGYBILL question several times before a missed payment ever shows up in a credit file, if it shows up at all.
For policymakers and regulators, the divergence underscores the need to treat utility insecurity as a core affordability issue rather than a niche credit-market problem. Relying solely on credit-bureau trends risks missing the millions of households who juggle bills, borrow from relatives, or cut back on other necessities to avoid shutoffs. Conversely, the Census data cannot reveal how often temporary hardship hardens into long-term debt that damages credit scores and limits access to future housing or loans.
Bridging that gap will likely require better data-sharing agreements between utilities, state regulators, and federal statistical agencies, along with privacy safeguards. Linking anonymized utility arrears to neighborhood-level Census responses could show where energy assistance, weatherization programs, or rate-design reforms would have the greatest impact. Until then, the one-in-six figure from the Household Pulse Survey should be read as a conservative measure of severe distress in a country where far more people are still one high bill away from falling behind.
