OK, file this away under the category, “Things that drive me nuts.”
By Wolf Richter for WOLF STREET.
The Census Bureau and the Department of Housing and Urban Development jointly produce the data on new single-family houses: permits, construction starts and completions, inventories at all stages of construction, sales at all stages of construction, median and average contract sales prices, and related data.
Whatever happened during the pandemic in collecting and processing the data, the Census Bureau and HUD decided that the pricing data was totally screwed up and needed to be fixed, and they dramatically revised the sales prices that had been collected during the pandemic.
And today, as part of this, they announced huge revisions to the pricing data going back through 2020. For example, they chopped off $36,000 from the median price at the peak in October 2022, taking it from the old $496,800 to the new-and-improved $460,300.
We look at the three-month moving average because it irons out much of the month-to-month squiggling. The revisions chopped off $38,000 from the three-month moving average of the median price at the peak, taking it from $480,000 to $442,000.
In other words, the revisions unwound $38,000, or 25%, of that $150,000 pandemic-era spike. The spike, based on the old prices went from $330,000 in April 2020, to $480,000 in October 2022.
There were three phases of these revisions:
- April 2020 through January 2023: All median prices were revised down.
- February 2023 through December 2023: Median prices were not revised at all.
- January 2024 through March 2024: All median prices were revised up.
This chart shows the old median prices through March (blue line) and the revised median prices through April (red line), as a three-month moving average.
For your amusement, here is the chart with the monthly data, not three-month moving averages:
The announcement.
The Census Bureau, when it announced the revisions today, provided some details, including the five items below. Note #4:
- “The sales price range groups in Table 2, ‘New Privately‐Owned Houses Sold, by Sales Price’ have been updated to better reflect the current distribution of new home prices.
- “New price groupings have also been introduced in our time series file ‘New Houses Sold and For Sale by Price Range.’
- “Data between January 2020 and March 2024 have been re‐calculated incorporating any additional data and revisions received since initial publication and re‐released in the new price groupings.
- “All tables containing historical median and average sales price data have been revised between January 2020 and March 2024.
- “With this release, seasonally adjusted estimates of housing units sold, housing units for sale, and the months’ supply of new housing have been revised back to January 2019.
Data collection was totally screwed up during the pandemic?
This single-family residential data is collected as part of the broader Survey of Construction. The Census Bureau explains the details in the Methodology. It starts with construction permits that are then followed by field staff, including visits to the permit offices. A sample of projects are then followed through various stages of construction, to determine starts, completions, and sales. This data is obtained from builders via surveys:
“The Census field representatives use interviewing software on laptop computers to collect the data. Facsimiles of the computer-based questionnaires are provided to respondents to familiarize them with the survey. These facsimiles show the questions that are asked for housing units in single-family buildings on Form SOC-QI/SF.1 and in multifamily buildings on Form SOC-QI/MF.1….
Then the data has to be compiled and extrapolated to the overall US economy.
Now imagine doing all this in 2020 and on, when people were working from home, or pretending to work from home, when travel and other activities were restricted, when all kinds of shortages and delays and huge cost increases drove the construction industry up the wall, so to speak….
Whatever happened, the Census Bureau and HUD decided that somehow the pricing data got totally screwed up and needed to be fixed, and they revised the figures that were collected during the pandemic. But the sales and inventory data of new single-family houses going back to 2020 were revised in relatively minor ways, compared to the pricing data.
Enjoy reading WOLF STREET and want to support it? You can donate. I appreciate it immensely. Click on the beer and iced-tea mug to find out how:
Would you like to be notified via email when WOLF STREET publishes a new article? Sign up here.
I wonder how this data compares to industry data such as Zilllow or Case-Shiller.
Case-Shiller cannot include new-builds into its data because it only counts sales pairs, so houses have to be sold at least for the second time before they make it into the data.
It’s the we-don’t-know-what-we’re-doing-the-dog-ate-the-data-we’re-just-gonna-make-it-up revision.
The revised data, reducing past spikes down and pumping current numbers up, makes the housing market seem like it’s in a nice and stable “new normal” at the 30%+ higher price – rather than unstable and teetering on the edge.
Whew, lucky us! It would be terrible if anyone got the idea that housing is a major part of the biggest asset bubble of all time, ready to blow at any moment.
All provided by the same experts that we are expected to trust as much now as we were expected to in the past, when the previous “wrong” data was published.
On the brighter side, when the bubble bursts, if the “emergency” is of a lower magnitude (starting from a lower high), maybe the emergency response interventions will be commensurately smaller?
One can hope…
Also an easy way to convince first time home buyers to agree to government funds/tax credits to “buy” at a low price, only to be indebted to the same government for 60 years….I.E. student Lon’s.
So they are basically lying? Not surprised by this at all.
MW: Nvidia’s stock surge at odds with the S&P 500 — to a degree not seen in 8 years
NVDA 9.32%
Nvidia sees $221.2 billion boost to its market cap on the day
Nvidia is sporting growth one tech CEO says hasn’t been seen ‘in the history of capitalism’
Wheeee!!! Way to go! /s
I remember the day I sold it at $120 a share before any stock splits and I thought it was high at the time.
I went to the casino and had my chips on red and won
After that I wanted to put it all on 36 but did not do it. It was 36.
Did I loose my fictitious winning or did I not risk loosing all I had?
Starting to see more and more stories about the paucity of mega profitable current AI “use cases”.
The enemy of hype is the little kid pointing out that the Emperor is waltzing around with his ******* hanging out.
The amazing part is that it is only like three years since people were paint huffing about the Metaverse.
When was the last time you heard about the Metaverse?
I predict the same fate for AI. Within five years, it will be a has-been technology.
It already is way past has-been and never will be.
It isn’t so much that I hate on AI projects per se, it is that they are so hyper-hyped beyond their current actual utility by Wall Street hucksters desperate for the next huck-to-stir (“Keep-the-Grossly-Inflated-PEs-Flying!”) and the most moronic media clickbait tramps (“World Altering Invention That Does I Do Not Know What!”).
Cas…agree it’s that seemingly eternal (and often-profitably exploited) human desire to believe in magic, made so-much easier when presented in a thin, brightly-printed wrapper of (often-pseudo) techno-science…
may we all find a better day.
Agree with Dustoff 110% as usual, but since this article is about things that drive Wolf nuts, WHY do people (eg, CAS) now say “hate ON” instead of just “hate”. Has “on” become the verb and hate the adverb? (or whatever modifies “on”?) Does it imply anything different or is it just “cool”?
Could one say, “why are you being mean ON me” and would the message be different than “mean TO me?
No, I don’t know why I think about shit like this…..senile….?
NBay – have come to the conclusion that Prof. ‘enry ‘iggins’ “…cold-blooded murder of the English tongue…”, thanks to our modern communications technology, has developed into more of a mass-casualty situation…(not excusing myself, at that. Grin, bear it, and best!).
may we all find a better day.
Thanks, I was just going to assign it to usual suspects, Black ghetto word games, LA, or just the social media players…mostly kids……LOL, etc. Have to look up this Higgins character.
Yeah, mass casualty and plenty wounded and confused. Doesn’t really bother me, though, just curious, and wandering out of box for no reason as usual.
Later.
Wolf, I wonder what new home construction would look like in 2024 California if it wasn’t mandated from Sacramento? I’d wager there’s a large delta in there, somewhere.
Ted – ‘…just add water…’?
may we all find a better day.
A better methodology such as dividing the sales price by the square footage including garage, followed by multiplying that number times the assessors square footage might even show a price drop.
Still a crazy price spike even with the revision.
Hmm,
In the land of Oz these figures would be used by local council to set rates.
Credit due?
I have no words, no words. This is a bit surreal! Thanks for sharing WR.
You said it. And any words I’m thinking right now are probably better left unsaid.
“The most splendid housing bubble”
“the most splendid housing revision”
“The most splendid 4 year moving average”
The most splendid reporting…
Thanks Census Bureau for making homes bit affordable for common Joe.
/s
Bottomline: One can’t trust these govt manipulated metrics.
This kind of thing really makes you wonder. I am becoming very skeptical of what my government is telling me.
I’ve been skeptical of what my government told me since the 70’s.
In 2016, then CEO of statistics Canada- Wayne Smith – suddenly resigned from his position. The reason cited was statistics Canada’s loss of independence. I remember reading about it when it happened and just thinking, this isn’t good. How can people not be skeptical and how can such federal agencies retain credibility?
It’s like a bird with half of a beak missing, not a good look for attracting a mate.
Future glances back will never know the fine beak that was.
The average American is hurting..
or are they?!
Just revise figures down when it comes to various coats and bam there you go. Not so bad now.
Wolf, I was thinking that now you have the opportunity to update all of your charts. Interesting to see your approach.
I look at all revisions. Most of the time, they’re too small to worry about. The three-month moving average, which includes revisions, normally takes care of it. Sometimes when the revisions are big and go back years, as today, I make a big deal out of it. I have done that periodically for a long time.
YES,,, EXACTLY!!!
And just one more reason I send you money every year.
YOU actually EARN every dollar, OK, hundred,,, that I send…
Many thanks again for your GREAT analyses and especially the WOLF’S wonder reporting,,,
VVNv
Just another area for AI to possibly trim the fat. All these overlapping government jobs with folks tripping over each other to perform the same function and still can’t get it right. I smell a change in the air coming in November, “ Left turn Clyde “ might be coming to the end of the road. Little Gomer Powell now says he doesn’t know if Fed rates are having any effect on inflation, sometimes you just have to shut your pie whole, but that would defeat the Shock and Awe dog whistle to Wall Street. Keep hiking or start lowering, or just STFU until something happens.
pie hole red one
Never forget – AI totally depends upon the data it is “fed”. Data collection, cleaning, normalization, etc., is a world unto itself.
Always has been.
Or build better models.
“But the sales and inventory data of new single-family houses going back to 2020 were revised in relatively minor ways, compared to the pricing data.”
If data collection was messed up from the pandemic wouldn’t all of those items be adjusted about the same magnitude? This smells like they just want to avoid showing any price declines in housing. Doesn’t inspire confidence in the “all govt stats are lies” crowd I bet.
Math/Accounting Problem:
If you went to the supermarket with a $1 coupon and spent $5, they charge you $5 and hand you back a $1 for the coupon, what is the price?
According to generally accepted accounting practice, the price would reflect the discount and would be $4.
In the US housing market, the price would be $5, you would be taxed on a $5 house, and the $1 “concession” would go to a lender for a buy down, or lower your closing costs, but the price would never be reduced.
Cookdoggie,
“…wouldn’t all of those items be adjusted about the same magnitude?”
No, that’s an illogical connection. Counting something is one thing, coming up with an accurate nationwide median price is a very different thing. That’s an issue in ALL housing data. The counts are easier to establish; the pricing data is all over the place.
I get tired of people making up conspiracy theories everywhere because they don’t understand the complexity of reality. This stuff is just dumb. It’s like the ancient Greeks that made up the story that a guy/god named Helios was driving a chariot across the sky every day because they had no clue about how the universe worked.
Buy why try to come up with a nationwide median price? Prices are dependent on location. The USA maybe not as much as the EU. In the EU an EU median price would be nonsense, due to the variaton between locations.
“In the EU an EU median price would be nonsense, due to the variation between locations.”
Nonsense or not, there is a home price index for the EU, released quarterly by Eurostat. I’m going to start a series called, “The Most Splendid Housing Bubbles in Europe,” with data from Eurostat. In addition to price indexes for each country, they have price indexes for the entire EU and for the Eurozone.
In the US we have several country-wide home price indexes as well, nonsense or not. Two days ago, I reported on this one:
https://wolfstreet.com/2024/05/22/home-sales-whacked-by-mortgage-rates-active-listings-price-reductions-jump-to-highest-in-years-but-sales-of-high-end-homes-surge/
The Greeks thought it was a chariot, but we know now that the sun is being towed across the sky by a flying saucer. We’re just so smarter than they were back then! /s
David in Texas,
Thanks for clearing up my misunderstanding. I thought it was the new 600hp Ford Raptor pickup truck that was towing the sun across the sky these days.
LIKE the analogy Wolf:
Also like the attempt, however unlikely to convince those who consider Every effort to make sense of the VAST delta between ”statistics” and personal experiences.
“Never the twain will meet.” Comes to mind, frequently.
Hi Wolf,
I can’t wait for your first article in the Europe’s Most Magnificent Housing Bubbles series to come out.
Waiting for the Q1 data for all EU countries to be released.
@Cookdoggie I’m not in the “all govt stats are lies” crowd, I’m on the “govt. will manipulate data to get what they want crowd”. Like @Malthus I worked with polling and surveys in college and found that 99% of the time the people paying for the poll or survey (they don’t pay for themselves) want a specific answer ahead of time and the polling companies know this and can phrase questions to get the answers they want (“Do you support the current mayor who has done so much to make the city better or his challenger who plans to raise taxes and fire cops and teachers”… P.S. I knew Econ Majors who were sad to find out that getting a job as an “economist” meant you had to lie for a living (Google National Association of Realtors Chief Economist Liar for an example)…
“I’m on the “govt. will manipulate data to get what they want crowd”.
LOL. I’m on the “real estate investors and brokers will manipulate data to get what they want” crowd.
The rest of your comment is manipulative BS because you’re comparing small-scale political polls of 1,000 people (at best) to these massive multiple-source data collections that included data from permitting offices, field staff visiting projects, and surveys of builders (which I explained in the article). You’re FAR worse than the government in your manipulative efforts.
Dang Wolf, if I wasn’t already on the toilet, I would have definitely crapped myself after reading your last sentence! I felt that one.
Cookdoggie—
“Doesn’t inspire confidence in the “all govt stats are lies” crowd I bet.”
It’s not so much that ALL government stats are lies, but that all the the stats are presented as irrefutable facts, rather than often being tainted with bias, ineptitude, and intent to manipulate and control. Hayek, in his Nobel prize speech, called it “a pretense of knowledge.”
These stats are considered facts until they aren’t. The old facts are revealed as new ignorance. By then, though, policy is already in place which benefits some, damages others, and nearly always includes unforeseen consequences.
And so it goes.
I wish I could “revise” last year’s taxes owed. That would be rather convenient for me.
I remember ‘income averaging’ in the old days, then they did away with it.
As 44-year broker, new home data has been flawed more than just recently – it was completely warped in RTC days ’89 to ’95 and ’08 GFC dramas. Similarly on resales, Realtors & C-S only report what was in MLS and teen percentages of trades are never in MLS – gotta go to tax records for true action where some county jurisdictions are prompt, while others lag months behind. TFS
This data here, as explained in the article, is NOT based on the MLS. It’s based on permits and projects and builder surveys.
The Case-Shiller index, which we cover here, uses public records data for prices (except in Texas), not the MLS.
The NAR data, which we also cover here, is based on MLS data.
New home sales numbers are further warped as most (assume all) trades are packed with closing cost & financing juice. TFS
No, these are based on contract prices often months before the close (at the time of the sale and the contract price, construction on the house may not have even started).
Which causes another problem: that deals that don’t close stay in the data, which is normally not a big thing because it is roughly a constant, but it was a huge thing in late 2022, when 30% to 50% of the deals were canceled because they were overpriced as mortgage rates began to soar, and so they never happened, but the pricing data from those canceled overpriced deals wasn’t removed. This may be where part of the revisions came from.
Here is one of the articles about those cancellations – written in November 2022 when the October data was released. The cancellations got worse still in Nov and Dec 2022:
Massive Cancellations Make Mess of Already Low New-House Sales. Inventory Glut at Deep Housing Bust 1 Level. Buyer Traffic Plunges
Click on the chart to enlarge:
Pretending to work from home is greatly preferred to pretending to work in the office!
“If you can’t hit the target, move the target until…….”
Cheers,
B
The revision doesn’t sound believable to me.
So given the reduction, this means property tax appraisals will be reduced by all the counties as well and we’ll be receiving a big fat property tax refund check in the mail, right? /s
I wonder if the people who bought in late ’22 or early ’23 who are probably still upside down if they had to sell quickly can go back and “revise” what they paid and get the seller to return fifty or eighty grand to them? Show them this revised chart as evidence?
There is going to be a group that has those magical 3% mortgages but who also paid tippy -top dollar who won’t break even until 2030 or so.
Selling due to a job loss or other unanticipated situation for the next few years – there’s no equity there and they might even have to bring cash to the closing table to walk away.
In late 2022, cancellation rates hit 30% to 50% as reported by builders. The pricing data here is based on contract prices often months before the close (at the time of the sale and the contract price, construction on the house may not have even started). But deals that don’t close stay in the data, which is normally not a big thing because it is roughly a constant, but it was a huge thing in late 2022, when 30% to 50% of the deals were canceled because they were overpriced as mortgage rates began to soar, and so they never happened, but the pricing data from those canceled overpriced deals wasn’t removed. This may be where part of the revisions came from.
So it looks like the most overpriced deals of 2022 got canceled. Builders than sold the homes with reduced prices, and those deals closed.
Ah very interesting, thanks. So the sweetest deals at the highest prices with the lowest rate mortgages were more like late ’20-early ’22.
Why did it take nearly four years to discover and report the problem?
Election is coming …
Biden forgave more student loans
Biden released more oil from spr.
I don’t endorse either candidates
Fueled up the work truck today.
Diesel is 20 cents/gallon cheaper than
Regular.
Been alot of years in my area since that has happened.
“With this release, seasonally adjusted estimates of housing units sold, housing units for sale, and the months’ supply of new housing have been revised back to January 2019.”
Are they adjusting past estimates to reflect current estimates, or are they adjusting past estimates to reflect ACTUAL activity?
If the statistics are never trued up to actual, I view that as a problem.
I’m reminded of the battle of the bulge when an America brigade was offered surrender or die by the German O.I.C. and the American CO responded with “Nuts”. Ah,the battle among central banks over empire.
Interesting stuff Wolf. Does this data feed into inflation data? For example are the CPI or PCE going to adjusted down for the last couple of years and slightly up for this year?
No, but that would make a fun article 🤣❤️
It is simple, in one sense: Either the reported sales prices were wrong, or the sample of sales used was not representative.
This begs the question: Which error mechanism was dominant? If sales price reported in error, how is that possible? If the sampling was biased, what flaw in the methodology caused that problem?
And what did the Census Bureau do to correct the errors? It sounds like CB did not go back and re-collect/fix the original data, but rather only “adjusted” the final median to their liking. Which would be a quite questionable action at best.
Finally, who benefits from adjusting/revising the price data down? Is it the Fed and Congress, who created the insane housing inflation? The Fed would benefit in two ways, by being able to say “The peak prices were not as high” and “the drop is not really that big”, as others have pointed out with other words. Who else is behind the revisionism?
Finally, I agree that basing the price index on self-reporting from builders is a terrible methodology to begin with. Pricing data should be based on recorded sales prices. The local taxman rarely will miss finding the price that a house really sold for, I think.
I should add that I just saw the Wolf comment that one error mechanism is that pricing data was reported by builders according to contracted prices, on deals that may not ended with a closed sale.
I’m not saying that the cancellations are the reason for this screw-up, I don’t know, but they must have played a big role in it, that’s my guess.
Here is one of the articles about those cancellations – published in November 2022 when the October data was released. The cancellations got worse still in Nov and Dec 2022:
Massive Cancellations Make Mess of Already Low New-House Sales. Inventory Glut at Deep Housing Bust 1 Level. Buyer Traffic Plunges
Click on the chart to enlarge:
Wolf,
One thing I have notice a lot is that the sales contract gets modified to add in additional seller concessions (meaning seller pays for termite remediation, home warranty, home repairs, etc.).
Since these are non-cash to buyer concessions, the sales price remains the same.
Normally, these expenses are small, but I have seen cases where they are $20-50K, which has a real impact on the data.
And they are much more common now than I have ever seen before.
Do you have a sense of how prevalent this is, and how much it screws up the data?
Plus there are now the mortgage-rate buydowns that are costly to the builder, but are not a price reduction of the contract. Each individual builder can obviously track all their data. But for outsiders to collect accurate costs of across-the-country purchases of new houses seems like a huge challenge because there are a bunch of big factors that are not part of the contract price. In theory, they can get this cost data from the builders via the surveys.
“…It’s BIGGER than Kong! BIGGER than Gojira! It’s GIGO! KING of the monsters! Now playing in theaters everywhere!…”.
may we all find a better day.
Cant let a good pandemic go to waste, let’s screw the numbers up, down all around.
Let’s have a third party, a fourth party, a party hearty look into the matter…see if a rat is at play.
What do you think of those apples?
[Content deleted by Wolf]
Hey SocalJim,
You can promote anything legal on this site, including real estate, cryptos, etc., as long as you do it properly:
By buying an ad.
But you cannot promote real estate or whatever in the comments for free. That’s just cheapskate stuff.
The place for your future promos:
Look on the homepage…
https://wolfstreet.com/
…see the gray box below the third headline? That is an ad space that I sell direct to advertisers. It’s a “sponsored post.” You write the article, “by XYZ,” or whatever, and you can promote and link anything legal, including your own business, housing going to the moon, your collection of classic motorcycles you’re trying to sell, or whatever. Check out the sponsored post that is there now to get a feel for what you could do. Think about the possibilities!!!
You can rent this space for $500 per 7-day period, prepaid at the time the unit goes live. I haven’t raised my price since I started offering this service two years ago = 0% services inflation. This is really cheap – almost free – to promote whatever. You can contact me via the email address under the “Contact” tab, and we can work out the details. Cheers!!
This is great. If only other sites deleted their RE shill (and crypto shill, and all other shill) comments and forced them to buy ads the web would be a much better place. ZH should’ve been doing this for the last 5 years at least. I have no idea if it’s still as bad because I don’t go there anymore — too many shills.
Thumbs totally UP CSH!
Joke.
No. Advertising and promo is not free on this site. The current ad in the gray box on the homepage (ends on Monday) promotes a bitcoin mining stock. I’d be happy to work out a deal with you to promote something else, such as housing. I have no idea why real estate promotors are such cheapskates, always expecting to get free promos and ads.
Crazy Uncle on Facebook: The government is lying about all the statistics!
Woofstreet reader: The government is too incompetent to figure out what the real statistics are in the first place
Hanlon’s Razor:
“Never attribute to malice that which is adequately explained by stupidity.”
I never, never, never trust data from China! Oh, wait….
Data errors happen all the time as Wolf mentioned with the issues with Covid. Sometimes these take time to either appear or something changes that creates an opportunity for revisions. I have seen plenty of errors that were not discovered in private industry that can cost millions to correct and I’m sure there are many more that either are not caught or just not revised as long as the financials are not impacted . Nothing illegal about revisions 🎉
Wolf,
I wanted to get your perspective on a discussion I had with a prominent real estate developer (primarily apartment complexes).
He said he thought we might be near, or even in, a recession already. To him, the present feels a lot like mid to late 2007.
He also echoed what you are talking about with the unreliability of the data, and that banks are pulling back from financing deals because of concerns about comparable sales pricing data, valuations, etc.
His take was that the Pandemic, stimulus, etc. had so distorted the economic data that it wouldn’t surprise him if none of the economists saw it coming until it hit.
Granted, he is operating in one narrow corner of the economy, and one narrow geographic area, but I thought this was a very eye opening discussion.
If you’re a CRE developer, you’re in a depression right now, not a recession. But that’s not the whole economy, just a slice of it.
As an old guy who was in the middle of the Greatest surge in RE devs, circa 2007, I agree that now is very much ”like it was” then…
Caveat buyer, etc., is just the beginning IMO…
Statistics can be interpreted many ways. Years ago the US and Russia were in a world athletic competition – the US placed first and Russia placed second.
The US newspapers reported it as US first place, Russia second place.
The Russia newspapers reported Russia second place, US next to last.
I always wait for the revisions after an administration change before I take any actions. Or at least until the 5 year anniversary.
It’s all getting so tiresome. Can we get a revisions index, similar to a volitivity index, on government statistics?
Government statistics (propaganda) are just a political tool and American politics is just high school for geezers.
BTW, the revised numbers included the entire last year of the Trump administration and then continued with the Biden administration, and was revised under the Biden administration. So I don’t see how partisan politics played into this.
In the comments above, I showed a chart about the massive cancellations in late 2022, when the price peak occurred, and I linked my article from that time. I think that provides a pretty good shot of what might have happened to the pricing data, if 30% to 50% of the contracts are cancelled, but it’s the contract prices that are reflected in the data. So check out those comments and the link.
Seeing everything in terms of politics clouds your vision. Lots of crappy data collection happened during covid because everything was screwed up. The private sector made an endless series of bad decisions because they didn’t know either what was going on, and we’ve documented some of them here. It was just an awful time for doing anything correctly.
the Census Bureau still has field agents driving from permit office to permit office to collect data on their laptops…in rural non-permit areas, the agents drive the backroads looking for signs of construction activity, and do on-site interviews when they find it
Changing historical data makes accurate analysis useless except for creating a fantasy. Whether it is changed to create positive or negative is somewhat immaterial. It is still misleading.
These figures from the Census Bureau are only relevant to the extent that denominator (cost of construction and real estate, new single family home) remains stable. In fact, we all know that the large home builders are maintaining/growing margin by reducing the square footage of homes they build, building them on smaller lots, and reducing the cost of the finishes (windows, roofing, siding, flooring, tile, fixtures, etc). It may well be that they can hold selling prices stable – by selling cheaper homes. They will do almost anything to avoid discounting the selling price (on paper) – and have the tools to buy down mortgage rates and/or offer “free” up-grades so that they are effectively discounting – without showing a lower sales price.
If one purchased a home during the revised time frame, will a rebate check be coming?