So Caltrain is cutting the middle of the day service (see press release: http://www.caltrain.com/news_2009_07_02_board_approves_changes.html). Which is reasonable, given that it's pretty empty, but also the wrong direction from the ridership growth standpoint.
Fundamentally, we can lay the blame for this on municipalities allowing office parks -- if the jobs were by the train line, the train would have all-day demand.
Tuesday, July 7, 2009
Friday, January 30, 2009
How to get the cost of living under control, part 1b
It's a basic truism of real estate that in a given area, rents (or mortgages) tend to all fall within the same range on a per-square foot basis. That is, the rent on a 2000 square foot house tends to be about 4 times the rent on a 500 square foot apartment. This is not strict, of course -- yards, views, maintenance, etc. can all have a large effect on price -- but it tends to hold true overall.
Or to put this another way, the claim that "houses cost more because people want houses" is false. Houses typically cost more than apartments because houses enclose more square footage than apartments. A 2000 square foot apartment is likely to cost as much as (or more than) a 2000 square foot house.
Or to put this another way, the claim that "houses cost more because people want houses" is false. Houses typically cost more than apartments because houses enclose more square footage than apartments. A 2000 square foot apartment is likely to cost as much as (or more than) a 2000 square foot house.
Ultimately, the three components of cost in new residential development are the same for single- and multi- family housing: cost of land, cost of design/legal approval and cost of construction. For most developments, the cost of construction is far and away the biggest cost. In a given range of quality, cost is basically proportional to square footage enclosed, though taller buildings tend to cost more per square foot to construct (because each additional story calls for heavier materials, etc.)
But in more desirable areas, the cost of land becomes a bigger component of the overall cost, and as cost of land grows, it makes sense to use less land on a given residence. Accordingly, desirable area such as cores of cities have more apartment buildings and fewer houses.
Since the old downtowns of Valley cities are generally desirable areas, you'd expect to see lots of apartment buildings near the Valley's old downtowns, and to some degree, you do. But you don't see nearly as many as you might expect given the land costs, largely because the Valley's policies increase the cost of design and legal approval and shape developments into less efficient forms.
Take College Avenue, just north of the California Ave business district in Palo Alto. It's mostly zoned for medium-density multi-family residential (RM-30). It has a mix of low-density apartments buildings and single-family housing now, so at first glance, you'd expect building higher-density housing could be a good investment for a developer. And since the street is a short distance from both Stanford campus and Stanford Research Park and near a Caltrain stop, the California Ave business district is doing well even in the economic downturn, and the Palo Alto schools are among the tops in the nation.
But even with high property values, Palo Alto's codes don't leave room for developer profit. RM-30 zones require 60% of a lot to be open space, don't allow building above 35 feet, and require a maximum floor area ratio of .6 to 1. So assume a developer could buy two houses next door to each other, for a total 10000 square foot lot. Said developer could build a total of 6000 square feet of building, or about enough for 5 two-bedroom apartments or 8 one-bedroom apartments. Note that the "higher-density" housing we're talking about here is barely increasing density over an existing two houses -- even if both of the existing houses were no larger than the 3-bedroom bungalows typical of the neighborhood, we're talking about 2-4 extra bedrooms, total. The replaced houses wouldn't be cheap to start with and if we're trying to increase affordability, a developer would need to be able to sell or rent to more people to spread out the costs, so not being able to add many units bodes badly.
Back to the costs: According to the code, a building with 5 two-bedroom apartments needs to provide a minimum of 12 parking spaces. A building with 8 one-bedroom apartments would need to provide a minimum of 14 spaces(!). Parking spaces are big: the typical estimation is that a space plus its share of access is 325 square feet. 14 parking spaces would take up 4550 square feet, or basically half the double lot. This typically means the developer would need to bury the parking underground. Underground parking is very expensive -- even with efficient layouts, underground parking tends to cost on the order of $30,000-$40,000/space, and an underground garage for a mere 14 spaces probably won't be even that efficient. Since the cost of the 14 spaces would need to be spread across the 8 one-bedroom apartments, each apartment is looking to cost nearly $60,000 (or about $350/month in rent) for parking alone.
The excessive parking requirement is just the biggest of the many ways Palo Alto makes costs increase: Palo Alto requires design review on all projects, limits heights, requires all street trees to be fenced and preserved, etc. Some of these decisions may be justifiable (I love Palo Alto's street trees, personally), but they do increase costs.
So the situation we have today is this: we need new construction near old downtowns to get the cost of living under control, but that construction itself costs too much, in part because of policy decisions increasing costs. To improve the situation, so we need to make policy changes to make construction near old downtowns affordable. I'll leave it to another post to detail the specific changes.
Sunday, January 25, 2009
How to get cost of living under control, part 1a: More housing in old downtowns
First, we have to change zoning to encourage more housing near the commercial centers, especially in the historical town centers.
Look at it this way: Rents in Silicon Valley are high. Rents in San Francisco are a little higher. But the overall cost of living is much higher in Silicon Valley than in San Francisco because most Silicon Valleyers have to drive and most San Franciscans don't.
This is not because Muni (SF's bus system) provides the mobility that a car does. Buses are slow and crowded, fall off schedule and don't always go directly where you want to go. But quality of life doesn't demand mobility for its own sake; quality of life demands access to goods and services. And San Francisco has good commercial neighborhoods, so a San Franciscan can get groceries, toiletries, coffee, etc. without needing to drive. And in the cases where a San Franciscan *does* need to go across town to get to work, see a friend, etc., Muni is often adequate. And it's much, much cheaper than owning a car.
Which doesn't mean that San Franciscans don't own cars -- there are about 400,000 cars registered in San Francisco, or about one per household. But that's one fewer car per household than in Silicon Valley. And at about $6000/car/year (once you add up depreciation, insurance, gas, capital cost, repairs, parking, taxes, tickets, costs below insurance thresholds, etc.), SF's overall cost of living drops well below Silicon Valley's once you take the cost of transportation into account.
In theory, of course, car ownership is a matter of choice. Certain free-market economists might argue that Silicon Valley families own more cars than San Franciscans just because they choose to, and that economist wouldn't be wrong. But it's misleading to frame the issue of car ownership as one of personal choice unaffected by environment -- it's just not possible to access a diversity of goods and services in Silicon Valley without a car, and that fact strongly influences Silicon Valley families' choices. The decision balances quality of life and cost of living -- when comparing San Francisco and Silicon Valley, a similar quality of life costs more in Silicon Valley.
So coming back to the point of lowering the cost of living without sacrificing quality of life: one good way to do that is to reduce car dependency. And one good way to do that is to focus new development around existing walkable retail areas.
Conveniently, many valley cities have healthy retail in their historical downtowns (University and California in Palo Alto, Castro St. in Mountain View, Downtown Redwood City, etc.). These historical downtowns areas are the perfect locations for increasing housing density without increasing car dependency -- the areas have walkable retail, good transit connections (almost all the downtowns are there because the train line is there) and good bike connections (and they're all flat, too).
So why don't they have more housing already? Here's where we get into zoning rules, construction economics, urban design and inter-city competition. In other words, another post.
Look at it this way: Rents in Silicon Valley are high. Rents in San Francisco are a little higher. But the overall cost of living is much higher in Silicon Valley than in San Francisco because most Silicon Valleyers have to drive and most San Franciscans don't.
This is not because Muni (SF's bus system) provides the mobility that a car does. Buses are slow and crowded, fall off schedule and don't always go directly where you want to go. But quality of life doesn't demand mobility for its own sake; quality of life demands access to goods and services. And San Francisco has good commercial neighborhoods, so a San Franciscan can get groceries, toiletries, coffee, etc. without needing to drive. And in the cases where a San Franciscan *does* need to go across town to get to work, see a friend, etc., Muni is often adequate. And it's much, much cheaper than owning a car.
Which doesn't mean that San Franciscans don't own cars -- there are about 400,000 cars registered in San Francisco, or about one per household. But that's one fewer car per household than in Silicon Valley. And at about $6000/car/year (once you add up depreciation, insurance, gas, capital cost, repairs, parking, taxes, tickets, costs below insurance thresholds, etc.), SF's overall cost of living drops well below Silicon Valley's once you take the cost of transportation into account.
In theory, of course, car ownership is a matter of choice. Certain free-market economists might argue that Silicon Valley families own more cars than San Franciscans just because they choose to, and that economist wouldn't be wrong. But it's misleading to frame the issue of car ownership as one of personal choice unaffected by environment -- it's just not possible to access a diversity of goods and services in Silicon Valley without a car, and that fact strongly influences Silicon Valley families' choices. The decision balances quality of life and cost of living -- when comparing San Francisco and Silicon Valley, a similar quality of life costs more in Silicon Valley.
So coming back to the point of lowering the cost of living without sacrificing quality of life: one good way to do that is to reduce car dependency. And one good way to do that is to focus new development around existing walkable retail areas.
Conveniently, many valley cities have healthy retail in their historical downtowns (University and California in Palo Alto, Castro St. in Mountain View, Downtown Redwood City, etc.). These historical downtowns areas are the perfect locations for increasing housing density without increasing car dependency -- the areas have walkable retail, good transit connections (almost all the downtowns are there because the train line is there) and good bike connections (and they're all flat, too).
So why don't they have more housing already? Here's where we get into zoning rules, construction economics, urban design and inter-city competition. In other words, another post.
Thursday, January 15, 2009
Cost of Living, Livability
There have been some news stories recently about people leaving California, primarily as a result of high unemployment and a high cost of living in the state. Of course, Californians going elsewhere is an old trend -- the Pacific Northwest and Intermountain West have been complaining about incoming Californians for decades now.
But still, it's easy to understand why people leave, especially when the economy is bad and the cost of living in California is as high as it is. And this is especially relevant to Silicon Valley because the cost of living is especially high here.
Which brings up a question: why does it cost so much to live in Silicon Valley? The Econ 101 answer is simple: demand is high and supply is low. But that answer brings up the Econ 102 (or maybe Poli Sci 101) question: if demand is so high, why is supply still low? There are a lot of factors, but one major component is a misguided sense of what it takes for a community to be livable.
Silicon Valley policies seem to be designed to maximize cost of living: lots of single-family housing (= higher land costs on a per-person basis, both because each person needs more land and because more land is taken out of the supply for housing), jobs distributed across the region (= car-oriented development, which necessitates more car ownership and more land for roads), high parking requirements (= more land needed for businesses, and thus higher land costs for business) and lots of land protected from development (= less land supply, and thus higher land costs for developable land).
Of course, if we tried to change any of these policies, people would object in the name of livability: Upzoning single-family neighborhoods would bring complaints that quiet, safe neighborhoods would be "ruined" by apartments. Attempts to centralize jobs by prohibiting new commercial development outside of core areas would be seen as increasing commute times for those in the suburbs (and un-American, too). Decreasing parking requirements in neighborhoods would be charged with killing cute neighborhood retail by "forcing" customers to bland regional shopping malls (with parking). And opening the foothills or baylands to development would decrease open space, which seems to be necessary for sanity in urban America.
But if we don't make some changes, all this livability will continue to cause people not to live here.
Coming up: What we need to do to get costs under control without losing anything that actually makes the region a pleasure to live in.
But still, it's easy to understand why people leave, especially when the economy is bad and the cost of living in California is as high as it is. And this is especially relevant to Silicon Valley because the cost of living is especially high here.
Which brings up a question: why does it cost so much to live in Silicon Valley? The Econ 101 answer is simple: demand is high and supply is low. But that answer brings up the Econ 102 (or maybe Poli Sci 101) question: if demand is so high, why is supply still low? There are a lot of factors, but one major component is a misguided sense of what it takes for a community to be livable.
Silicon Valley policies seem to be designed to maximize cost of living: lots of single-family housing (= higher land costs on a per-person basis, both because each person needs more land and because more land is taken out of the supply for housing), jobs distributed across the region (= car-oriented development, which necessitates more car ownership and more land for roads), high parking requirements (= more land needed for businesses, and thus higher land costs for business) and lots of land protected from development (= less land supply, and thus higher land costs for developable land).
Of course, if we tried to change any of these policies, people would object in the name of livability: Upzoning single-family neighborhoods would bring complaints that quiet, safe neighborhoods would be "ruined" by apartments. Attempts to centralize jobs by prohibiting new commercial development outside of core areas would be seen as increasing commute times for those in the suburbs (and un-American, too). Decreasing parking requirements in neighborhoods would be charged with killing cute neighborhood retail by "forcing" customers to bland regional shopping malls (with parking). And opening the foothills or baylands to development would decrease open space, which seems to be necessary for sanity in urban America.
But if we don't make some changes, all this livability will continue to cause people not to live here.
Coming up: What we need to do to get costs under control without losing anything that actually makes the region a pleasure to live in.
Tuesday, January 13, 2009
Geography of peninsula cities
Peninsula cities are slightly different from each other, but they all tend to have some things in common: Office parks near 101, an old downtown linking a Caltrain stop and El Camino, strip malls along the rest of El Camino, upper-class housing in the hills, and (closer to) middle-class housing in the flatter part of town.
This geography reflects some history: El Camino came first, as a stagecoach route, and the downtowns were founded around stagecoach stops. The Southern Pacific came next (on the tracks that now house Caltrain), extending the downtowns between the old stagecoach stop and the new train stop. As the towns grew, they added the common housing of the time, often 2-3 bedrooms, often bungalows. When the interstate era came along, ubiquitous cars opened new areas for development: the whole strip along which the stagecoach had run now had access, but it all needed parking: El Camino became car-oriented retail. Meanwhile, housing sprawled into the hills and the lowland spaces between towns.
Why does this matter? Because when and why a building was built usually defines its form, and its form usually defines its later function. Downtown Palo Alto has more in common with Downtown Mountain View than either has with the office parks filling in the space along 101.
This geography reflects some history: El Camino came first, as a stagecoach route, and the downtowns were founded around stagecoach stops. The Southern Pacific came next (on the tracks that now house Caltrain), extending the downtowns between the old stagecoach stop and the new train stop. As the towns grew, they added the common housing of the time, often 2-3 bedrooms, often bungalows. When the interstate era came along, ubiquitous cars opened new areas for development: the whole strip along which the stagecoach had run now had access, but it all needed parking: El Camino became car-oriented retail. Meanwhile, housing sprawled into the hills and the lowland spaces between towns.
Why does this matter? Because when and why a building was built usually defines its form, and its form usually defines its later function. Downtown Palo Alto has more in common with Downtown Mountain View than either has with the office parks filling in the space along 101.
First Post
This blog is intended to discuss trends in architecture and urban design in Silicon Valley.
Subscribe to:
Posts (Atom)