Pressure of Truth
Exposing the spin on all sides of the news.
Finance

Consumer Confidence Index Reads 89.4 in August; New-Home Sales Ran at a 607,000 Annual Rate in July

The Conference Board's index slipped 0.8 points to its lowest level since January, while the Census Bureau put July new-home sales 10.5% below June — a change the agency flagged with a margin of error wide enough to include no change at all.

How spun is the coverage?Coverage bias 3.3 / 10
5 sides analyzed15 sources cited

Two Numbers, Same Morning, Opposite Signals

On August 25, 2026, two federal-adjacent reports landed within hours of each other, and both got read as bad news. The Conference Board said its Consumer Confidence Index fell to 89.4 in August, down 0.8 points from a revised 90.2 in July[1]. Separately, the Census Bureau and HUD reported new single-family home sales running at a seasonally adjusted annual rate of 607,000 in July, 10.5% below June[2].

Forecasters had expected confidence to land closer to 90.2 or 90.3, so the miss was real, if small[3]. The housing number was the weakest pace since January[7]. Both figures fit neatly into a narrative of an economy losing steam, 70 days before the midterm elections[4][11].

But the Conference Board's own headline on its release reads "US Consumer Confidence Edged Down Slightly in August[1]." That's not spin from a defensive source. It's a clue that the number everyone quoted hides two other numbers moving in opposite directions.

The index that fell is actually two indexes, and they disagree

Consumer confidence isn't one measurement. It's built from two separate surveys stitched into a single headline number. One asks how people rate business and job conditions right now. The other asks how they expect things to look six months out.

In August, those two halves split hard. The Present Situation Index — today's conditions — rose 6.8 points to 121.2, its first gain in four months, after three straight declines[1]. The Expectations Index — the six-month outlook — fell 5.8 points to 68.2[1].

That second number carries a specific warning attached to it. The Conference Board says readings under 80 on the Expectations Index have historically preceded recessions[1]. It's a pattern from past cycles, not a forecast, but it's the kind of detail that gives each side of the political argument something to point at.

Because the two halves moved in opposite directions, both camps can quote the same release honestly and land in different places. One side gets to say conditions today are fine. The other gets to say the outlook is flashing a warning light. Neither is misreading the data — they're just each reading half of it.

A margin of error wide enough to erase the headline

The housing number needs its own translation. Census reports new-home sales from a sample of transactions, not a full count, so every monthly change comes with a built-in margin of error. July's reported 10.5% drop carries a margin of ±14.0%[2].

Do the arithmetic on that and the honest range runs from a 24.5% plunge to a 3.5% increase[2]. The agency isn't hiding this — it prints the figure in its own release. It's just rarely the number anyone quotes.

There's a second wrinkle in how the 607,000 figure gets built. "Seasonally adjusted annual rate" doesn't mean 607,000 homes sold in July. It means July's actual sales pace, adjusted for the time of year, then scaled up as if the whole year ran at that speed[2]. A modest real shortfall in one month can turn into a large-looking annualized swing.

None of that makes the number meaningless. It does mean a single month's move, especially one this close to the margin, is a shakier foundation for a big claim than most headlines let on.

Why the same 0.8-point dip reads as a warning to one side and a distraction to the other

Breitbart's headline was blunt: "Consumer Confidence Drops As Midterms Loom[4]." Its argument leans on the Conference Board's own finding that confidence weakened among Republicans and independents while it improved among Democrats, which it reads as voters bracing for a possible change of control in Congress[4]. It also points to the Present Situation gain as evidence that conditions on the ground are actually fine.

Center-left and mainstream outlets read the same release the other way. Newsweek paired the confidence drop with President Trump's approval rating and a "70 days before midterm" countdown[11]. U.S. News tied the Expectations decline to tariffs, the Iran conflict, and prices that haven't come down[13].

Both sides have a real incentive at stake, not just a talking point. The administration and its allies want to keep the economy from becoming the defining midterm issue, and separating "today" from "six months from now" helps that case[4][11]. Democrats and center-left commentators want cost of living to be the frame voters carry into November, and the sub-80 Expectations reading gives them a concrete number to attach to it[1][13].

The Conference Board itself sits outside that fight. Its business is credibility with corporate members and financial markets, and overstating a move costs it more than understating one — which is likely why its own release chose the word "slightly[1]."

The unsold homes stacking up regardless of who's right about the mood

Set the political argument aside and the housing sector has a plainer problem: too much unsold inventory. There were 488,000 new homes for sale at the end of July, equal to 9.6 months of supply at that month's sales pace[2]. Four to six months has historically counted as a balanced market, so builders are sitting well past that[2][6].

They're already responding on price. The median new home sold for $393,800 in July, the lowest since July 2021, and 63% of builders offered incentives in August while 35% cut prices outright[2][5]. HousingWire, which covers the industry, frames this as a buyer-favorable market rather than a crisis — the falling prices are the story's point, not evidence of a collapse[5][6].

Yet buyers still aren't showing up in the numbers that matter. Contract signings fell roughly 13% year-over-year in the South and roughly 43% in the Midwest, that region's weakest stretch since 2012[5].

The reason ties back to a mechanic that both political camps are downstream of. A mortgage rate compounds over 30 years, so a small rate move changes a buyer's monthly payment more than a similar-sized price cut does. The 30-year fixed rate averaged about 6.54% in July, the highest since August 2025[5][10]. A cheaper house at a higher rate can cost a buyer the same each month as a pricier house did a year ago — which is why builders cutting prices to a five-year low hasn't been enough to bring buyers back on its own[14][15].

What the coverage left out, and who benefits from leaving it out

Lay the outlets side by side and the tells are visible in the headlines themselves. Bloomberg went with "US New-Home Sales Decline to Lowest Level Since January" — a datestamped comparison with no adjective attached[9]. Continuum Economics, a UK macro research firm writing for institutional clients rather than a domestic political audience, put "mixed detail" directly in its headline and named no political actor at all[7].

Quartz's "seven-month low" is technically accurate and also the most alarming true thing you could say about a move of less than a full point, with the Present Situation gain left out of the headline[8]. Breitbart's numbers check out too, but its causal story — that the Expectations drop reflects fear of a Democratic Congress — is its own inference rather than something the survey directly tested[4].

The detail almost nobody leads with is the Census Bureau's own ±14.0% margin sitting next to its 10.5% headline decline[2]. It's easy to see why: naming it weakens the alarmed framing and the reassuring one in equal measure, which means neither side benefits from putting it up top. What happens to both the mortgage rate and the Expectations Index between now and November is still an open question, and it's the Federal Reserve, not either party, that holds the lever most directly tied to it[7].

Like this article?

Share this article

The Bias Ledger average rating 3.3

The same story, as framed by outlets across the spectrum, ordered least to most biased. The bias score (1 = straight, 10 = heavily spun) is an AI assessment of that framing — click an outlet to see its track record. The tell is the word choice or omission that reveals the angle.

OutletVantageBiasHow they frame itThe tell
BloombergU.S. center, financial-market audience1"US New-Home Sales Decline to Lowest Level Since January" — a datestamped comparison, no adjective.Cleanest headline in the set: it says how far back you have to go rather than characterizing the move. Written for readers who will check the margin of error themselves.
Continuum EconomicsUK-based independent macro research, subscription-funded1"U.S. July New Home Sales, August Consumer Confidence — Both weakest since January, mixed detail."Puts "mixed detail" in the headline itself, which is the single most accurate two-word summary available. No political actor is named anywhere. The limitation is the reverse: it never asks why the outlook component is falling.
The Conference BoardU.S. business-membership research group, funded by corporate members — not neutral, but with a reputational stake in not overstating moves2"US Consumer Confidence Edged Down Slightly in August" — with the Present Situation gain named in the second sentence.The word "slightly" and the equal billing given to the 6.8-point rise in current conditions. This is the source everyone else is summarizing, and almost no downstream headline keeps the qualifier.
HousingWireU.S. mortgage and real-estate trade press; audience is lenders and builders3"New home demand swoons even as sales prices hit a 5-year low" and "Builders face a tougher math problem as completed inventory rises."The framing is inventory-first and politics-free — the villain is 9.6 months of supply, not any official. That is genuine analytical value, but the audience is the industry, so builder margin pressure gets more sympathy than buyer affordability.
QuartzU.S. center-left business site4"U.S. consumer confidence hits seven-month low in August 2026" — the true-but-steepest framing of a 0.8-point move."Seven-month low" is accurate and also the most alarming true thing you can say about a decline of less than one point. The Present Situation gain is not in the headline.
Breitbart NewsU.S. right6"Consumer Confidence Drops As Midterms Loom" — the fall is real, and its cause is the prospect of a Democratic Congress.It reports the data accurately, including the downward revision to July, then supplies a causal story the survey does not test. The Conference Board's release reports that confidence weakened among Republicans and independents while improving among Democrats, but the survey does not ask respondents about control of Congress — the midterm-fear explanation is Breitbart's inference, not the source's.
NewsweekU.S. center-left6"Trump Approval Rating as Consumer Confidence Drops 70 Days Before Midterm" — the economic reading is packaged as a political scoreboard.Pairing an index number with an approval rating implies a link the article does not establish. The countdown framing ("70 Days Before Midterm") does no reporting work; it points the reader at an election forecast.

References

  1. US Consumer Confidence Edged Down Slightly in August — The Conference Board · U.S. business-membership research organization funded by corporate members; publisher of the index
  2. Monthly New Residential Sales, July 2026 — U.S. Census Bureau and U.S. Department of Housing and Urban Development · U.S. federal statistical agency; primary source
  3. US Conference Board August consumer confidence 89.4 vs 90.2 expected — InvestingLive · Commercial financial-markets news site, trader audience
  4. Consumer Confidence Drops As Midterms Loom — Breitbart News · U.S. right, explicitly conservative advocacy-oriented outlet
  5. New home demand swoons even as sales prices hit a 5-year low — HousingWire · U.S. housing and mortgage trade publication; advertiser-funded, industry readership
  6. Builders face a tougher math problem as completed inventory rises — HousingWire · U.S. housing and mortgage trade publication; advertiser-funded, industry readership
  7. U.S. July New Home Sales, August Consumer Confidence — Both weakest since January, mixed detail — Continuum Economics · UK-based independent macroeconomic research firm; subscription-funded, institutional clients
  8. U.S. consumer confidence hits seven-month low in August 2026 — Quartz · U.S. center-left business news site
  9. US New-Home Sales Decline to Lowest Level Since January — Bloomberg · U.S. center, financial-data company; subscription and terminal revenue
  10. New Home Sales Drop to Six-Month Low in July — The Real Deal · U.S. commercial real-estate trade publication
  11. Trump Approval Rating as Consumer Confidence Drops 70 Days Before Midterm — Newsweek · U.S. center-left, traffic-driven digital news
  12. US Consumer Confidence (topic page and August 2026 release detail) — The Conference Board · U.S. business-membership research organization; publisher of the index
  13. Consumers See Concern About the Future of the Economy — U.S. News & World Report · U.S. center to center-left general-interest news and rankings publisher
  14. New-Home Sales Tumble, Giving Buyers More Leverage With Builders — National Mortgage Professional · U.S. mortgage-industry trade publication
  15. New-home sales tumble as lower prices fail to bring buyers back — ConsumerAffairs · U.S. consumer-facing commercial review and news site