The latest (third quarter) personal income stats from the Bureau of Economic Analysis came out today. Washington State showed a 0.6% increase from the second quarter to the third, but a record 1.5% year-over-year drop from the third quarter 2008.
I thought it would be interesting to compare this measure to home prices. Keep in mind that what we’re looking at here is statewide income and King/Snohomish/Pierce home prices, so it’s not a perfect comparison. For this chart I have indexed the BEA data to Q1 1990 = 100, and re-indexed the Case-Shiller index to January 1990 = 100.
This is a similar comparison to one Deejayoh made in June, but the statewide data is released quarterly (vs. yearly for county data), so we can get a more timely snapshot. When we view the same type of x-y scatterplot that Deejayoh posted, we can see the reversion to the long-term trend as well. The following chart is based on the raw BEA and Case-Shiller data without any re-indexing.
Home prices are still slightly above where they would be if their 1990-2000 relationship between incomes had held steady, with the latest Case-Shiller index reading coming in around 12% higher than the trendline. This falls roughly inline with my expectation that Seattle-area home prices probably have about another 10% left to fall before they reach a natural market equilibrium.
When It Comes to Unemployment and Household Income Statistics
The BLS should be renamed the Bureau of Loose Statistics or Baloney Labor Statistics….LOL
As long as the spiked increases in household incomes of the elite are averaged in with the flat or decreases of everyone else, let-alone more workers per household [whether it be more working living in a house or one working two jobs] due to the chronic recession/depression we’re in….the BLS data is a good change comparison YOY or month to month base source measuring stick, but a lousy base for house qualifying indicators for the middle incomes.
Most of already know the BLS unemployment data omits the U6 severely underemployed data as unimportant….tell that to those households as they apply for food stamps…LOL
Is the X axis labled correctly, or is my cold medicine more powerful than I thought? I don’t get it.
RE: Scotsman @ 2 – Oops sorry, that’s in millions of dollars. I’ll add that.
Good stuff Tim. The X-axis is total personal income for the state? unit is Millions?
RE: deejayoh @ 4 – Assuming we’re talking about the 2nd chart, stolen from you, that’s correct. That is the statewide total personal income number from the BEA, in millions.
Makes sense. But since the bubble caused the natural equilibrium to go so far out of whack, I think it’s possible or even likely that we’ll see greater than another 10% drop.
Also, how is the personal income data derived?
If, for example, they are counting ten people, one whom is a CEO, five who work at McDonalds, and four who work at Wal -Mart:
If the CEO got a big bonus for getting TARP money, bailout money, and laying off all his employees, but everyone else stayed the same, doesn’t that make it appear as the personal income on average has increased?
Does the small minority of very highly paid folks skew the numbers?
In regards to the scatter plot:
Starting with the x-axis, aggregate income increased from:
~$70,000 M in 1990
to
~$275,000 M in 2007
Some years had larger increases than others, and the dot com bubble shows up around 2001, but income trended up for 17 years. Starting in ~2007, income stopped increasing.
Now moving on to the y-axis:
Concentrating especially on the 2001 to 2007 period, but it appears that I could start at ~1995, housing prices only went up. (The chart has a bit of a resolution problem to be 100% certain, but with just a bit of smoothing, housing prices always increased from 1990 to 2007)
If incomes and prices were directly correlated, then we would expect that the shift near the end would not be present. In other words, the data points would have remained near the historic data. Since incomes have remained fairly constant over the last two years, if the relationship held, then all the data points would be very close to each other. In this case, “close” is measured from a best fit regression from 1990 to 2007, or include all data points if that suits your fancy.
Note how this is not the first time that Seattle has had a period of stagnant incomes. Incomes were flat from ~2001 to ~2003 (dot com) . What’s interesting, however, is that while incomes remained flat during that period, housing prices continued to go up, albeit at a slower pace than during periods where incomes increased.
Now let’s look at aggregate income starting in 2005 at about $250,000 million. Roughly speaking, the income increased from 2005 to 2008, albeit that the period from 2007 to 2009 is basically flat. The data is a bit rough in this area, and we are not provided with error bands.
Today the CS HPI is about 150. This is the same as it was about 5 years ago, but aggregate incomes have changed from approximately $225,000 M to $275,000 M.
This makes me wonder what has happened to the population, especially in terms of people, but to a lesser extent in terms of homes. If the human population has increased at a higher rate than incomes, then that could be a factor in pushing the CS HPI lower. Similarly for the number of homes–as the quantity of homes goes up for a given static population, then there would be a downward force on the CS HPI.
As always, all of this must be balanced against the game of the greater fool.
One discussion point is: how reliable is your assumption that the income/price ratio must follow the same trend as between 1990-2000?
There might be other interesting variables to look at. For example, unemployment/price ratio or (size of Seattle population)/price. Also things like interest rates or the difficulty or easiness to borrow money could play a role.
RE: Ira Sacharoff @ 6 –
According to Pete Murphy, In His Book “Five Short Blasts”, Chapter 1, If The Economy’s Doing So Great, states in part:
“…This means that the top one or two percent of families are doing extremely well, driving most of the growth in the mean income of the entire population!…”
Ira, I just read the book, it was written in 2007, so its far worse today than it was then….BTW, the book would make a great Christmas present….LOL….
There’s little to no relashionship between median incomes and median home prices. It is simply statistical garbage.
The income disparity in the last 30 years wiped out even a small relashionship. 30% of bottom earners can afford nothing, they are not in a home buyer pool. Yet you use median, where they are included.
You can see the trend with just about any consistently used data, but your attempts to make some sort of scientific price prediction based on these charts are laughable.
The economy is terrible and financing is tight, thus you have downward pricing pressure. That’s it all there is to it.
By 2kt @ 10:
But, what we are trying to do here is to quantify things and try to create some clarity. Maybe there is a way to quantify this “financial tightness” you describe and then relate it to home prices.
Tim:
I revised and sent my article to you yesterday. It deals, in part, with a similar subject area as your article here. However, it uses Median Household Income instead of Personal Income. I believe household income is a more fair comparison, because many households have dual incomes. When comparing by individual incomes, things get unaccounted for.
You can see by this chart, comparing by Median Household Income shows a much wider divergence between incomes and home prices than 10%. There could be other reasons since I use a different source for median house prices than you do. However, the years throughout are relative to a single data source, so the ratios should be accurate compared to each other (comparing past ratio to present ratio).
http://housingcorrection.com/misc/MPtoMHI3.jpg
The data is good through Q2, 2009
By 2kt @ 10:
Do you suggest cutting the bottom 30% of earners out of the sample data? Have you considered that perhaps this sizable fraction of wage earners who are unable to buy a home is part of the overall phenomenon of home prices growing relative to wages?
By 2kt @ 10:
Aren’t you confusing the results of the housing bubble with the cause of the current “downward pressure” on (presumably) the prices of homes? I would reexamine my analysis if I were you.
Tim:
Looking back at the the previous thread, it looks like the difference is less than I suppose. I haven’t figured out the reason for the difference yet. I’ll keep looking.
A bit of a miss-match since the income data isn’t adjusted for inflation, but CS is (In a sense). As presented, it really doesn’t show the divergence of the bubble years.
RE: 2kt @ 10 – “There’s little to no relashionship between median incomes and median home prices. It is simply statistical garbage.”
This particular post deals with aggregate income, not median income.
RE: Scotsman @ 15 – Income isn’t adjusted for inflation? Some people define inflation by changes in income.
RE: fabuladocet @ 13 – The reason median income is irrelevant to home prices is many earners have no opportunity to buy a house, or possibly no desire to. For example, a collage student who works is part of the median income.
By Kary L. Krismer @ 18:
If you are arguing that the median income of home purchasers is different from the median income of all citizens, I agree. However, income to home price ratio is a long standing and meaningful factor in gauging affordability and the ability to qualify for a loan. It’s meaningful to compare past ratios with present ratios as long as changing household demographics are taken into consideration.
In a world where few people put 20% down, income to price ratio is especially meaningful because the majority of purchases are made with borrowed money based on the borrower’s income. People have been leveraging debt at much higher ratios than the past, which has resulted in a surface-level appearance that people have more money to spend than in the past. However, the consequences of this has been the massive number of late payments and loan defaults we are seeing now compared to the recent past. This clearly demonstrates that the overleveraging trend that helped fuel the housing bubble was not sustainable, and we are now witnessing a very necessary correction of this pattern.
It would be nice if the incomes the majority of people make in a particular community had no bearing on the price the citizens of that community can afford to pay for houses. But in order for that to be true, we would have to have a whole lot of people who make considerably more than the median. The demographic data shows this isn’t the case. The number of rich people is actually quite limited. And who would want to live in a world where most of a community’s citizens cannot afford to buy a house anyways? That sounds like a society where a class of downtrodden and impoverished slaves service an elite class of rich homeowners.
According to my data, inflation-adjusted incomes have declined over the past 10 years while inflation-adjusted house prices have shot through the roof. The biggest thing that has allowed this to happen is looser lending standards, low mortgage rates, and market psychology. These factors are being removed from the game, which means, home prices will be hard set to appreciate until continued adjustments between the recent divergence of house prices and incomes are through working themselves out.
I believe the last statement might not hold true if it weren’t for the extremely high unemployment rate where many of the jobs lost are permanently lost. I’m keeping in mind the unemployment rate is actually higher than the massive peak in the 1980’s. However the counting rules were changed since then in order to make politicians appear to be doing a better job.
I predict, we will trend downward toward a historically normal homeownership percentage. Eventually inflation, population growth, and job growth will put supply and demand back into balance. The question is, who wants to pass up this historically unprecedented opportunity to save a down payment? It’s not often that you can save money for several years and not have house price appreciation significantly eat into the down payment amount that was saved.
RE: Jonness @ 19 – “If you are arguing that the median income of home purchasers is different from the median income of all citizens, I agree.”
I’d like to know just how different these two numbers are, and just how significant that difference is. For starters, which one is higher?
“However, income to home price ratio is a long standing and meaningful factor in gauging affordability and the ability to qualify for a loan. It’s meaningful to compare past ratios with present ratios as long as changing household demographics are taken into consideration.”
Even if there is some difference between the median income of the entire population from the home purchasers, if the difference remains constant, then the data can be normalized.
By Jonness @ 19:
I’m not sure how long standing it’s been. I think it’s just a fairly recent creation of NAR to support arguments entirely different than what the arguments most people here would try to support. In actuality two different cities could have entirely different ratios of median income/median house price for reasons that have little to do with the health of their housing market. The age of their residents could be one such factor. Zoning another.
Also, no one uses median income to qualify someone for a loan. Individual income, yes. I’m not arguing income is irrelevant.
RE: Jonness @ 19 –
Historically the down payment of 20% has been for investment property. It will be that way again in a few years. 5% for a residential purchase, at a fair market price, adjusting for a 4% inflation is plenty of down payment for a family home, more than enough.
Banks should have been lending on the value of the asset. They stopped doing that in the 1990s and were only interested on the income from generating these mortgages. They invested that income to generate profits.
So, in my opinion, the borrowers strength, or ability to repay, is nothing in the big scheme of things. Banks don’t care, they don’t have to. They are too big to fail.
RE: Kary L. Krismer @ 21 – Let’s talk error.
What’s the difference in the median income between the total population and those who live in owner-occupied homes? Is the difference significant?
Since those who are in owner-occupied homes are also in the general population, what is the difference in the distribution of incomes?
FWIW, I just asked an elderly lady, she’s in her 90s, and she suggested that she didn’t make much money. She is a home owner.
By Kary L. Krismer @ 21:
I totally agree one area might have the same median income as another area, and house prices could be higher in one of these areas than the other. Thus, comparing one community to the other is not an apples to apples comparison.
IMO, what matters is comparing a particular city’s past income to price ratio to the current income to price and adjusting for known demographic changes, such as an increase in the number of dual income households. This certainly is not a perfect comparison, but when looking back over a short period of time, such as 10 years ago compared to now, it’s going to provide a useful comparison.
Where we appear to disagree is I believe the income the majority of people earn in a particular area is relevant to the cost of the majority of homes in that area. You believe, and I apologize if I’m misinterpreting you, the only incomes that matter to the majority cost of homes in a particular area is the precise subset of people who choose to buy homes. I don’t disagree that this provides a precise ratio. However, the homeowership percentage shows the majority of people own homes. Thus, I believe the incomes of the majority of residents strongly correlate to the incomes of the majority of home prices and provide a useful comparison through time. To further improve the precision of this comparison, one could include the income percentile breakdown of the area in the past compared to current and make further adjustments.
If median house price in an area doubles over a 5-year period, and median income only goes up 10%, most likely this pattern is not sustainable. IOW, the meaningful subset of home buyers you cite begins to shrink as affordability declines. This is because the incomes of this subset of people are directly tied to the median income of the community they live within.
LOL! I wasn’t aware the majority of home prices could have incomes. Maybe during the bubble that was true though?
I suppose that should be:
Thus, I believe the incomes of the majority of residents strongly correlate to the majority of home prices [in that area] and provide a useful comparison through time.
Hmmm… a bivariate model that does not account for population growth or density. Use of a linear trend line despite visual indicaton that a positive 2nd order (upward curved) relationship between CSI & income exists in the 1990-1999 data. If anything, the second graph makes today’s prices look right in line with historical trends… and possibly below them.
This market might go down another 10%, but it won’t be because of some sort of “natural market equilibrium.” It will be because of the economy and the excess of foreclosures. Personally, I think in further drops in price will be “overselling” due to temporarily adverse economic conditions. 2010 will be a good year to buy. The recovery in prices should begin in late 2010 and accelerate into 2012-2013.
RE: softwarengineer @ 9 – okay now go read “In Defense of Global Capitalism” or any book that counters your book’s claims. You remind me of a child. You read one book that states one opionin and think you have discovered the one and only answer to what is extreamly complex and impossible to do more that guess about.
By mydquin @ 26:
My data looks different from Tim’s. Actually, the Seattle data is close, although mine shows more variance., By looking at a wider view of the area, the effects of the bubble and the current overpricing are more apparent.
http://housingcorrection.com/misc/waaffordability.jpg
My data sources:
Median household income: http://www.ofm.wa.gov/economy/default.asp
Median house price: http://www.ihsglobalinsight.com/Highlight/HighlightDetail2350.htm
Holy cow! I just noticed OFM updated its median household income data on its website. Seattle houses are even less affordable than my charts and recent analysis show. Apparently, local incomes are dropping like a rock compared to their previous projections. This is nothing short of astounding.
http://www.ofm.wa.gov/economy/hhinc/default.asp
Here are the old numbers for King County (as used in my analysis):
King County Median Household Income (OFM Oct 2008):
2006: 65,845
2007 prelim est: 68,152
2008 projected: 68,832
2009 my est: 68,832
Here are the updated revised OFM numbers (OFM Oct 2009):
2006: 66,936
2007: 65,981
2008 prelim est: 64,714
2009 projected: 62,810
In addition, Global Insight just updated its Median Home Price data for Q3. My analysis used Q2 data, so I’ll have to update in order to bring things more current.
BTW, I definitely agree with Tim, house prices will fall as a result of being overpriced compared to incomes. The bubble is bursting. Just as in Japan, this is a long drawn out process.