Falling housing prices in Oakland and elsewhere

So the current San Francisco Business Times has an article about how it’s rough times for developers in Oakland, in which we learn that resale condo prices have dropped 12% since last year, from an average price of $475,614 per unit to $418,901 per unit. One blogger at Curbed SF appears to think that this crisis is a result of overbuilding in downtown under the Brown era. Reading the story, my immediate thought was that it didn’t seem like 12% was really all that bad, given the current state of the housing market. So it got me wondering, what kind of drop in price did single family homes see over the same period?

So I started looking at the market profiles over at Altos Research, where I learned that the median price of a single family home in Oakland is $290,925. As for the decline in that price since last year, well, it’s a whole hell of a lot more than 12%:

It seems that condos are actually holding their value much better than single family homes in Oakland.

Then, while I was there, I decided to look at the fluctuation in home prices for our neighbors, since I had found the charts so interesting when David over at Brooklyn Avenue posted them earlier this summer. So FYI, in case you’re wondering how the mortgage crisis has impacted the housing market in other East Bay cities, here you go:

Alameda: Median single family home price: $697,889

Albany: Median single family home price: $633,875

Berkeley: Median single family home price: $812,607

Concord: Median single family home price: $394,132

El Cerrito: Median single family home price: $606,159

Emeryville: Median single family home price: $533,929

Fremont: Median single family home price: $662,087

Hayward: Median single family home price: $373,096

Piedmont: Median single family home price: $1,655,321

Pleasanton: Median single family home price: $901,008


Related posts:

51 thoughts on “Falling housing prices in Oakland and elsewhere

  1. dto510

    These data show condos in Oakland are now significantly more valuable than single-family homes, and have held their price better. Seems like developing lots of condos is actually helping city housing prices – if anything, there’s a glut of old-fashioned low-density housing with very little demand for it.

  2. dto510

    Also, Oakland is by far the most affordable city in the region. Clearly “affordability” policies don’t work as well as the market.

  3. justin

    I’d look at location, as well, as a factor in holding value, not just unit type. I’m sure housing of all types is holding value (or not losing as much value) if located in relatively desirable areas (Grand Lake, Rockridge, Temescal, downtown, Piedmont Ave)–areas that are walkable and transit friendly. As condos are disproportionately located in these areas, condo values on average would be higher. This just highlights the importance of densities, walkability and transit access.

  4. V Smoothe Post author

    Well yeah, obviously home values vary from neighborhood to neighborhood. Homes in nicer neighborhoods are going to hold their value better than homes in crappy neighborhoods, although I’d guess that safety would be a bigger factor there than transit access. I can post breakdowns of prices by zip code, not sure if I can get any more neighborhood specific than that. I’ll look into it.

  5. Max Allstadt

    Once again, I find myself staring at a glaring example of how arbitrary it is to judge anything based on municipal borders.

    Unless this data is available by zip code, or we know nothing. Even by zipcode might not be enough.

    Bear in mind that poorer neighborhoods, though they may lose value faster, they’ll make it up much faster too. When a neighborhood “turns” you can have 50% appreciation in a year.

  6. len raphael

    surprising that berkeley was doing as well or better than piedmont. but both show “flight to quality” which is to say that there are people with bucks willing to pay bubble prices if they get good schools and safety. piedmont is not walkable transit friendly.

    but temescal +1,000 sq ft single fam houses are not increasing like berkeley might be, but they’re briskly selling for low 600′s (eg manilla behind public housing project) high 700′s. lower rockridge the prices for the same is running high 800′s and up. schools are not good in either of those places for most kids. safety, is block by block. compared to sf victorian flats, prices in north oakland are still cheap.

    -len

  7. Robert

    Zip codes are nice, but may not provide much more meaning, unless you can get the 9-digit code info. My Zip code (94610) has dropped by only about 10-15% over the last couple of years, but it covers everything from parts of Piedmont and areas below 580 down to Adams Point, so what is that really telling me? 94607 covers JLS and parts of downtown all the way to West Oakland and the Army Base. Much too heterogeneous to give meaning. Also, the smaller you get, the noisier the data. Even single Zip Code data is pretty noisy.

  8. Art

    I actually don’t think Oakland zip code data is going to clarify much, except maybe in Rockridge–the zips are just too big. Piedmont Ave’s zip, for instance, also includes Montclair and the city of Piedmont; Temescal’s zip stretches west into Bushrod; etc. The other thing to bear in mind is the quality of the Oakland housing stock that’s on the market right now: it’s at the lowest level we’ve seen in decades, in part because people who bought investment properties or refinanced rentals are walking away. Check out all the “contractor’s specials” for sale…. And East Oakland is awash in REOs, so it’s not surprising that these constitute a huge chunk of the city’s sales. So your $290K median price is buying you a very different house (and different neighborhood) than your $450K median got you a year ago. That’s a challenge that cities like Albany and Alameda are a little better insulated from as small cities with higher average household incomes.

  9. dto510

    Len, Piedmont is pretty transit-friendly. Most Piedmonters live close to Grand Lake or Piedmont Ave, which have bus lines and walkable commerce. Justin also has a good point about condos being disproportionately located in transit-friendly neighborhoods. Since there aren’t so many condos in the hills, near the better schools, that suggests that transit and walkable ‘hoods are just as, or more, important than schools for housing prices. On the other hand, one could say that condo owners are less likely to have school-aged children and so schools are inherently less of an issue for condos, so that’s why they held their value better.

    I’d be interested in seeing zip codes for the parts of Berkeley bordering Oakland compared to those on the Oakland side. Piedmont doesn’t have its own zip code. And how about the Richmond area? Much of it is very middle-class, but it’s all the same school district.

  10. V Smoothe Post author

    I’m having a hard time understanding exactly what people are expecting from the market data. I mean, I think we can all agree that some neighborhoods are nicer than others, and that homes in nicer areas are more expensive than homes in crappy areas and likely hold their values better in a market downturn. That seems self-evident to me.

    What I’m interested in is how much values have dropped in Oakland in the last year, how home values in Oakland have held up in relation to our neighbors, and to a lesser extent, and how values have changed in different parts of Oakland itself. Such data on its own doesn’t prove anything, but I think that when looking at, say, 94618 (Rockridge) versus 94705 (Elmwood), it’s reasonable to use it to speculate in a very general sense about the influence crime and local government (taxes, schools) have on home values.

    What are the questions that would be answered by neighborhood-specific data?

  11. Max Allstadt

    All true. I am in 94612, but so is the Uptown. Houses around me are surely nosediving with all the crime of late, while people apparently exist who are dumb enough to rent a two bed at the uptown for $2100.

  12. avis

    I’m in 94606, which is full of great old Victorians and Bungalows, but the crime problem is getting worse here. My neighbor told me that she lost 100K in equity in the last 4 yrs-she paid $420 in 2004 and now her house is worth only $320. Can this be true?

  13. Max Allstadt

    DTO, Show me a two bedroom house in West O that rents for 2100, and I’ll show you more interior square footage than a two bed in the Uptown, at least one parking space included if not two or three, and a private yard.

    A 900 square foot two bedroom apartment in a two unit house in West O, with parking and shared yard: Probably under 1500, likely more like 12.

    I have friends at 33rd and telegraph who have a two bed in fine shape, with a recent kitchen and an enormous shared yard. They’ve been there for a little over a year and they’re paying a thousand. They have a woodburning stove for chrisakes.

    If you get on craigslist, and your patient, you can easily beat the prices at the uptown. On the otherhand, if you have the money to spend, and you don’t want to look around and find a gem, I’m sure the uptown has plenty of lovely amenities for the modern young urban professional.

  14. Art

    Avis, I think one of the challenges—and one of the questions V. is asking in looking at these data to begin with—is that it’s extremely difficult to tell how much equity a given home has lost until it’s sold. We can guess based on trends—and that’s certainly what the Assessor’s office is attempting to do as they reassess property—but I don’t think your neighbor can truly know how much her home is now worth unless it’s on the market and the offers she’s getting are for $320K. The tools that purport to give you current value (Zillow, etc.) are very, very rough gauges and are thrown completely by the sea of foreclosed properties that are skewing the numbers. (Oaklanders all know, for instance, that some neighborhoods vary in desirability or safety block-by-block or based on how close to the freeway or based on local school quality, but those tools take none of that into account–not to mention that they don’t consider the shape your home is in versus the high-end remodel down the street or the ramshackle house with a bad foundation two blocks the other way.) The only reliable way that I know of to guess at how much your home might be worth is to visit nearby houses that are for sale and see how they compare. (Is it a better floor plan? Is it a nicer place or bigger yard? Does it need a lot more work than your house does?) Then see what they ultimately sell for, and you’ll have an idea of your value today.

  15. V Smoothe Post author

    Max -

    I think it just really depends on what you’re looking for in a place. I mean, I’m constantly finding myself in complete shock at what people are willing to pay to live in West Oakland.

    I see 2 bedroom houses in Dogtown renting for over $2k/month, and my reaction is the exact opposite of yours – why on earth would anyone pay that to live in such a place, where you have to deal with the crime, the vagrants, take care of a yard and have no transit access when for the same price you could live in a huge 2 bedroom apartment near actual restaurants, bars, stores, etc., practically on top of BART and every bus line, with a pool, hot tub, and gym on site! Or I see my friends renting rooms in places that I consider barely livable for like $100/month less than I pay for my entire one-bedroom downtown apartment that I have all to myself. I just don’t get it at all.

    But then, I don’t have a car, am often out late at night, think yardwork is a pain, and don’t really have any use for any more space than I have. You obviously have different needs and different priorities. I don’t feel the need to make assumptions about your character based on where you choose to live, and likewise, I don’t really think it’s appropriate for you to basically call people who prefer a different living situation dumb and lazy.

  16. James H. Robinson

    In Oakland, zip code is the minimal level of granularity necessary. You almost need to do it block by block, like in San Francisco and other high-density cities. Also, the median price for a single-family home includes the foreclosures and short sales that are on the market now. Once the inventory of those types of homes diminishes, so will their affect on home prices.

    Also, take into account that home prices rose to unusually high levels during the early years of this century. It was like the dot-com boom, but for residential real estate. Basically, home prices are just falling to where they should be under normal conditions.

    That being said, I think Oakland has much potential. CEDA is putting money into redeveloping neighborhoods. Oakland is finally getting additional police. Hopefully, there will be a new mayor a couple of years from now. And those new condos downtown and uptown will eventually be occupied either with renters or buyers or both, which will bring a vitality (or at least a customer base) to that part of Oakland.

  17. Max Allstadt

    I really think that as usual, the statistics only bear out the macroeconomics. Every city is block by block and deal by deal. And the natural socio-geography (is that a word?) always defies being lumped into zip codes or municipal boundaries.

    The foreclosure stats are pretty nuts too, I gotta say. How many in 94607 this year?

  18. californio

    When there’s a gap as big as that between Alameda and Oakland, I tend to mistrust the data itself. I haven’t been in the market for several years, but in 2000 I almost bought a couple of apartment buildings in Alameda that were selling for about the same price per door as over here. Are SFR’s in Alameda worth over double those in Oakland at this point? There are a lot of crappy neighborhoods full of crackerbox homes in Alameda, too. Correct me if I’m wrong, but…

  19. Colin

    This is a pretty useful tool:
    http://www.zillow.com
    Type in a street name and it will show you every house that’s for sale now with an estimated price, and everything that’s sold within the last year. You can also do it by zip.

    It doesn’t break down the data into a graph of estimated value like Altos, but it does show you real value over a set period.

  20. Max Allstadt

    um according to that site, everything on my block is on sale, or sold within the last year. I don’t think that’s accurate.

  21. VivekB

    Whilst I would certainly be thrilled if Zillow’s estimated value of my house in Rockridge was accurate, my guess is that it’s at least 10-15% too high from what it would realistically sell for.

    Bummer too, i need to re-fi but have to hold off as i’ll get destroyed by the assessor; my neighbors are reporting valuations of $100K *under* what the realistic sales price, meaning Zillow could be 25-40% higher than what the bank will agree to valuing at.

  22. V Smoothe Post author

    For someone else’s thoughts on the issues with sites like Zillow and Trulia, see here and here.

    Generally, I think those sites are really designed for people looking to buy a home, or interested in the market in that sense – a really neighborhood based depiction of real estate. Personally, I’m not so interested in that because I’m not in the market right now, and also because there are no homes around where I live to buy. I’m interested in broad-based market data. Both neighborhood specific sites like Zillow/ Trulia and general information sites like Altos are useful, it just depends on what you’re looking to get out of the information.

  23. Robert

    VivekB –
    Unless you are trying to get your assessment reduced, refi will have no impact on your assessed valuation.

  24. VivekB

    Sorry, what I meant was “No way will I get the bank to agree my house is worth anywhere close to what Zillow says it’s worth”.

    I have a 1st and 2nd, and want to consolidate to just a 1st. But in this market, I may not hit the # I need to hit in order to do that right now given the wackiness in the credit markets.

  25. Steve Carney

    One reason condo prices have not dropped as much as home prices is b/c many of the developers of recent condo projects don’t want to sell condos at huge losses, so they either keep them on the market at prices above market value or lease them. I also talked to an agent for the 20th/Broadway Condominiums who said “We would sell a lot more if we dropped the prices, but we don’t want to mess up the comps for the people who already bought condos here… so we’re probably going to be leasing some out soon.” This was 6-9 months ago. Also, people who bought condos 5 years ago want a certain profit so they are much less likely to drop their price than someone who bought a single family house 15-30 years ago. THe foreclosure crisis is also driving home prices down because homeowners about to foreclose and banks who take over homes want to sell as quickly as possible. All this keeps condo prices artificially high — higher than their fair market values. So although condo prices might be falling at a slower rate relative to homes, one must also look at the NUMBER OF UNSOLD CONDO UNITS to get the true full picture.

  26. V Smoothe Post author

    Steve –

    As it says in the post, the prices in question are resale prices for condos – the price that existing units are actually selling for, not the asking prices for new units.

    Also, there are no condos at 20th and Broadway.

    I don’t really understand what you’re trying to say about “fair market value.” The fair market value is what people are willing to pay.

  27. Steve Carney

    Oops, meant to say Broadway/Grand. I will try to respond to the rest of your comment after work today!

  28. dto510

    Yeah, Steve, there’s no such thing as an artificially high price. Either a deal is closed or it’s not. This is documentation of market value – condos are now much more valuable than single-family homes, even though way more condos have been built than houses in the last few years. Sure goes against NIMBY claims of a “condo glut,” doesn’t it?

  29. Colin

    The interesting thing about zillow from my perspective isn’t their pricing estimates, it’s the listing of sale prices. Saves me a trip to the office of records.

    People were asking about specific neighborhood numbers – that’s why I posted that link. Ignore the estimated values – they are estimates, and based on a lot of factors that aren’t regionally specific and should be treated as a parlor game.

    Unlike V, I don’t find broad market profiles that interesting. I don’t think they serve as economic tea leaves or give much insight, especially with the same formulas being applied across all regions of the country. The algorithm used to compensate for density in Cleveland doesn’t work here, and vice versa. And while Altos doesn’t discuss their research models, they are most certainly not offering a true price median – mostly because that’s not very useful information.

    Even though I’m not looking for a house, the neighborhood-level view says more about what’s going on in the city to me. The numbers are concrete and I know the areas.

  30. Max Allstadt

    Condos are more valuable that existing single families, or new ones?

    DTO, what about the number of unsold condos. There is no such thing as an artificial price, at least not here. But there is such a thing as a price that people won’t pay. The consequences of having a lot of units in that category are worth taking a look at…How goes 10k?

  31. Colin

    I wouldn’t be so bold as to claim that condos in Oakland are more valuable than single family homes in Oakland based on a statistical modeling system that isn’t city-specific. Altos isn’t offering a true median.

    I’m not saying that it’s not true, I’m saying that this model is no basis for making that call with any certainty.

  32. dto510

    Max, we’re talking about resale values: there are far more existing condos than new condos, just like houses. Also like houses, new condos are generally more expensive than existing condos. The impact of unsold units (inventory) is taken into account by the market. If there are a lot of unsold newly-built condos, that should depress the price of existing condos. These data show that the 10k project maintains its relevance even in a down period, putting a product on the market for which there is great demand. We’re talking about a $100k premium for condos over houses, that can’t be explained away. Clearly there is a glut of single-family homes, not of condos. If there are more condos than houses on the market yet prices are higher, that is yet more indication of demand skewing toward density.

    Colin – why do you say that these data aren’t a true median or city-specific? They appear to be both to me. I guess we should look at the same source of data for both condos and houses, but sales data is pretty transparent (it’s publicly recorded) so that shouldn’t be an issue.

  33. Art

    DTO, what you’re forgetting is that there are many more single-family homes in Oakland than condo units, and the condos tend to be located in neighborhoods closer to the city center (downtown, Adams Point, Lake Merritt, JLS, etc.) That’s what you’re seeing a premium for, not the fact that it’s a condo. In any given neighborhood, most condos still sell for less per square foot than single-family homes of the same size. (Exceptions might be neighborhoods like West Oakland, where the newer condos offer luxury amenities and the houses may be much older and in poorer condition; condo developments are by and large much newer than most of Oakland’s single-family homes.) But at the heart of it, it’s a geographic thing. There’s a premium for living in safer, better-located neighborhoods. You’ll find a similar premium on single-family homes in those areas over single-family homes of the same size in, say, deep East Oakland.

    Here’s an interesting link: http://www.redfin.com/neighborhoods/13654/CA/Oakland
    They break down price per square foot based on both zip codes and individual neighborhoods. Note that the big chart seems to be based on list price, but if you click the individual neighborhood links, they’ll give you a breakdown of sale prices and also a chart of condos versus single-family homes for each area. Seems like this captures what people were asking about earlier.

  34. Max Allstadt

    I would posit that a huge part of the glut of single family homes has to do with the fact that there are so many in high crime areas. But I could be wrong. Anecdotally, what I see in East and West Oakland are miles and miles of single family detached.

    At the same time, the density argument that DTO points out makes a lot of sense. The younger, whiter, more urban-friendly generation that is buying first homes right now… It would make sense that they’d want density.

  35. Colin

    dto510 -

    From Altos’ website:
    “A lot of systems will provide you with an estimate of a home’s price using previously sold properties as sales comparables. While this number is nice to know, it lacks in a few critical areas. For example, the sample size of comparable sales is often too small (just a handful of homes that sold over a given time period) to draw any reasonable conclusions about what the market is doing right now. In a real estate market that may change rapidly, this information is insufficient to help you make informed decisions. ”

    “Our system analyzes properties listed on the market for sale. While we don’t disclose our proprietary models, we’d be glad to give you a little inside information to help you understand the how you can best use our research to your advantage.”

    http://www.altosresearch.com/altos/website/Faqs.page

    I’ve had some tangential work-related interactions with them, and while I don’t claim to be an expert, people I’ve worked for who are have gone into great detail about the sort of modeling that they and companies like Altos use. It’s complicated and interesting stuff. It’s also information that is targeted to specific markets, and how they break out the numbers reflects that.

    A true median is pretty useless because the basis for breaking out the numbers is highly subjective. Knowing the net total sales and value of those sales in a city is useless for any analytical statement other than “sales are up” or “sales are down”. Knowing, for example, that 100 houses sold at an average price of 200k this year, and 50 sold at an average of 400k last year doesn’t really give you any insight into the health of the housing market. All 100 of those houses could have sold for exactly 200k, and 1 of the 50 could have sold for 2 billion. Which market is healthier?

    So obviously you start breaking it down by categories such as single family home vs condo, or square footage. Clearly, not all single family homes are comparable, so you can start factoring in the neighborhood as well. This isn’t really about valuation of the house, it’s about how much it was sold for, so this might seem irrelevant, but the percentage shift of a house on Trestle Glen versus one on Foothill will be different, and that’s why these factors start coming in to play. In other words, in order to determine the percentage shift in sale price, you have to factor in where the sale happens. Prices around Grand Ave will fluctuate more with economic times and housing bubbles than they will on 90th. To get a true percentage of median shift you have to weight these factors.

    The ways in which you narrow that data to try to make it relevant are subject to a lot of pitfalls. The easiest example is how you adjust for outliers. In poor neighborhoods, houses are more likely to be “sold” to family for fractions of market value in order to keep property in the family. Allowing that sale to be factored in would greatly reduce the average sale price, and that’s not going to result in the kind of data you can sell to clients. You can’t just chop off the bottom end of sales, though, since you don’t actually know that that will compensate accurately. In general (and this is my understanding from comments made by someone in the biz of real estate-specific statistical generation as filtered through my graphic designer’s brain), you calculate a percentage of sales per bracket that are going to get factored out.

    There’s another outlier category of uninformed purchases that you have to compensate for. People move across the country for a job, for example, and buy a place site unseen based on whatever information they could gather. They almost always pay above market value.

    Both of those categories are factored out because they affect the sale price percentages inordinately. But since you’re using math to compensate for them instead of examining each individual home sale, you are introducing assumption and errors.

    Altos is a research company, and they try to break down the figures in ways that will be indicative future trends so that they can sell their data analysis. They don’t get clients by saying how many houses have sold, they get clients by convincing them that they can predict what’s going to happen. In other words, their interest is in value more than sale price, so these factors are even more pressing for them.

    There are a bazillion other factors that people try to compensate for, and how they do it always has an up side and a downside. If a house needs a new roof and it’s a buyer’s market they’re going to get at least 80% of that knocked off the sale price – how does that impact the final number? If you quantize percentage shift based on neighborhood, how do the urban pioneers in West Oakland factor into that? How about foreclosures, which can either be counted as a sale for $0 (the bank isn’t actually buying the property back, they’re just taking possession of it since the terms of the loan weren’t met), a sale for the bank price (the mortgage, minus a percentage of the interest that the home owner was paying), or no sale at all? (I don’t know much about how foreclosures are factored in, since my time being tangentially involved with real estate was at the beginning of this whole crisis, so if there was an accepted formula it wasn’t a factor)

    As far as being city-specific, Altos is a national company, and they depend on the same basic formulas to calculate all areas covered, then try to adjust them as best they can for local factors. There is no E=MC2 for real estate modeling, though, so their goal is to refine formulas over time. The formula for sales (let alone valuation) in Baltimore should be wildly different from the ones here, but they aren’t.

    So I’m not saying that it’s not true, I’m saying this isn’t proof one way or another because they aren’t presenting a true median, and even if they were it’d be a meaningless number. At best, I would say it’s possibly indicative. But the idea of condos holding their value that much better than houses doesn’t quite pass the smell test.

    Sorry to ramble.

  36. dto510

    Art – are there many more single-family homes than condos? That may not be true. I agree that condos are in good locations, that is one of the reasons they’re desirable. Since it’s not economical to build single-family homes in those neighborhoods, it’s one more reason why the condo boom is successful and helping prop up home values even in this downturn.

    Colin – as long as we’re comparing apples to apples, I don’t see any problem with drawing conclusions from these data. All sorts of factors influence the market. It’s still the market.

  37. Colin

    dto510 -

    We’re not comparing apples to apples, though. We’re comparing mathematically-weighted-estimates-of-apples-designed-to-provide-insights-into-future-markets to apples.

    Here is an example of a median* that gives a very different picture from the Altos Research median*:
    http://www.redfin.com/help/search/market-data#Condo:_$/Sq._Ft.

    According to this, price of condos per sq ft has gone way down and single family homes have gone way up. Both figures are right*.

    Unfortunately, these numbers start at the peak of the bubble. While it’s great for dramatic effect, it’s not necessarily the most accurate view of the situation.

  38. Max Allstadt

    DTO –

    is it economical to build single family homes in Oakland anywhere? I mean other than McMansions or real mansions up in more suburban style neighborhoods?

  39. Art

    DTO–Yes, many more single-family homes by a longshot. (Remember, we’re talking condos, not just multi-family units.) Very roughly, the 2006 American Community Survey estimates that of ~145,000 occupied homes in Oakland, just under half (67,500) are single-family detached residences. The remaining units are either single-family attached, mobile homes, or apartments, but only 15% (11,700) of those are owner-occupied. Assuming most condos are owner-occupied and the balance are multi-family rental buildings, condos and TICs make up a pretty small fraction of the total. I’m sure there are official numbers somewhere that reflect the actual number of condos in the city–this is just a quick guesstimate based on the easy-to-find stats.

  40. dto510

    Colin, the link you provide is a way to interpret house versus condo data but the graphs presented are examples, not actual data and certainly not about Oakland. The Altos data is weighted and I guess the Mark Co data is raw. However, though we do not have perfectly-matched data to compare, they’re pretty good, and the mathematical modeling issues you bring up only means that it’s not exact (house-price data is generally very good because sales prices are recorded and available publicly). The price differences between houses and condos are enormous. V is coming back with better data, but there is no way that this difference is only the result of an imperfectly matched data set. Also, anecdotally, these numbers make a lot of sense, and we’re got a lot of reasons above why condos in Oakland are more desirable than houses – these data do pass the smell test.

  41. dto510

    Max – outside the Hills, there are single-family home developments in outer East Oakland, and single homes are built on small lots throughout the city. Though I don’t think the General Plan forbids multifamily development anywhere but the Hills, so infill houses probably see their days numbered.

    Art – Thanks for the info, though it seems a little off to me (it doesn’t take into account recent condo construction, of course). The condo market is smaller than the house market, but it’s not so much smaller that it’s not statistically significant. Remember, many houses are rented or illegally occupied as multifamily, so not all the houses are part of the active market. None of this disproves the fact that condos are holding their value better than houses even though the supply of houses is limited and the supply of condos is growing. Therefore, there is more a house glut than condo glut, which is the point I was making at the beginning.

  42. Colin

    You’re right – I posted the wrong link.

    Here is data analysis for 2 neighborhoods in Oakland that give a per sqft price. These are sales medians* that disagree completely with the article posted:
    http://www.redfin.com/neighborhood/1130/CA/Oakland/Grand-Lake
    http://www.redfin.com/neighborhood/14/CA/Oakland/Adams-Point

    In the Grand Lake neighborhood, the house price per sqft price has gone down $50 since Oct 06, the condo price has gone down $200.
    In Adams Point, the house price per sqft price has gone up $100 since Oct 06, the condo price has gone down $150.

    Here’s all of 94610:
    http://www.redfin.com/zipcode/94610
    The house price per sqft price has gone down $50 since Oct 06, the condo price has gone down $230ish.

    Looking at the chart above the sqft charts (# of condos vs houses for sale) you’ll see why the median isn’t necessarily giving you the data you think you’re getting here.

    A sqft comparison is better than a per unit measure as far as I’m concerned. You could argue the opposite quite easily. We would both be right*. My point is it’s more complicated than you’re making it.

    The numbers do not pass the smell test, and I would suggest that the modeling is a big part of why. I’ll even go so far as to suggest they might not be based on actual sales data but rather projected sales data.

  43. Max Allstadt

    You’re right about the plan, dto, but lot sizes in the flatlands create plenty of situations where only a single family can be built. That’s why you see so many tiny empty lots, i think.

    I remember you recently mentioning that small scale development is particularly difficult to do in our city. I’d love to see a plan to fix that.

  44. dto510

    Colin, the RedFin dataset is miniscule. And it’s almost entirely condos. It’s not statistically significant. Am I reading this right that it’s based on fewer than 20 sales? The marketwide data seem far superior.

  45. Art

    Agreed that larger datasets are generally superior (honestly I don’t know why Redfin carves their data set up into so many neighborhoods–renders it almost useless!) But the city-wide data just tell you—correctly—is that if you had a portfolio of condos randomly selected across Oakland and a portfolio of randomly selected single-family homes across the city, your condo portfolio would be doing much better right now. You need the neighborhood-level data to understand how this transfers to an individual condo on the same block as an individual single-family home. I think you’ll find that in that situation, the two rise and fall roughly in tandem, and generally speaking the condo tracks slightly below the single family in terms of $/sf. Some Oakland neighborhoods are just holding their values much better than other neighborhoods, and many of the condo developments and condo conversions are in those stronger neighborhoods. (Doesn’t negate your point that a condo is a decent investment—I buy that.)

  46. Colin

    dto -

    It’s statistically insignificant why? Considering that the graphs show the ONLY statistical data available and the same data everyone else is using I think you’re mis-reading.

    The cart at the top shows the number of houses and condos sold in a given month, sorted by area. Looking at this chart:
    http://www.redfin.com/zipcode/94610
    We can see that in July08 (for example), about 30 houses were sold and 60 condos sold. The sqft value of those properties averaged at about $400 per for the houses and about $385 for the condos.

    The charts of sq ft pricing, while not on the same scale vertically, do show a decline in condo value by that measure.

    If you look at the summary data at the bottom (which is averaged over the entire time period graphed) you’ll see that condos stay on the market twice as long as houses, go on the market at $383 per sq ft (houses are $385) and sell for a median of $322 (houses are $351).

    The neighborhoods I liked to have a high percentage of condos, so I think it’s a more accurate comparison than all of Oakland – makes more sense than comparing condo at Broadway/Grand to houses down by the coliseum. You may disagree. Since there are more condos sold, outliers should have less effect on the average price, as should the relatively uniform date of construction. And the price goes down still sharply.

    Now: tell me why you think this data should be dismissed and the Altos data shouldn’t.