Monthly Archives: September 2015

You’ll Always Have the Antelope Valley

Purple City often provides a thought-provoking different viewpoint on how cities grow and what it means for a city to be available to all. Thus, you get support for the new bus system in Houston, but not for elimination of express bus routes when LRT lines open. Support for more urban density, but also for long-haul commuter rail lines and freeways that many transportation advocates don’t like.

However, I think a recent post misses the mark. It lists three reasons why lack of development in Palo Alto isn’t the problem in the Bay Area’s housing crisis: (i) land is scarce, (ii) the Bay Area is “blue tribe” – i.e. marrying and having kids later in life, and (iii) Silicon Valley attracts lots of smart people. The post argues that you don’t want to turn Silicon Valley into Singapore, because it would make humanity dumber by being a place that attracts lots of smart people but puts them in the kind of housing (small apartments in high-rises) that discourages them from having kids.

First, let’s address the assumption that this result would make humanity dumber. The snarky retort would be that watching TMZ would disabuse you of the idea that the ability to secure a position of wealth and influence is necessarily correlated with the possession of superior genetics. But at a more human level, this just isn’t how society works. I don’t claim to know the first thing about genetics, but regardless of the hand you are dealt by nature, the opportunities and resources society presents you with (or does not present you with) are far more important. Silicon Valley isn’t overwhelmingly white and male because of genetics.

Turning to the less controversial aspects, it’s true that California has a lower fertility rate than many “red tribe” states like Utah and Texas. However, the reasons behind these differences are probably quite complex, and I don’t think they can be boiled down to something as simple as being in a large SFR instead of a small apartment. The states at the bottom are a mixed bag of places with practically no one living in high rises, including Alabama, Pennsylvania, and all of New England. Meanwhile relatively dense Hawaii checks in near the top. In California, it is likely that the drop in the fertility rate has coincided with the decline in production of apartments, but it would seem questionable to draw any conclusions from that. To the extent there are differences within cities (young singles & elderly in apartments, families in the burbs), that may be self-selection.

What we can probably say with some certainty is that, all else equal, people will choose to have more children when they feel more economically secure, because raising children can be costly. High cost of housing is one thing that makes people feel less economically secure, as do things like potential medical and education expenses. If we are concerned about fertility rates being high enough to ensure a large enough number of future workers to pay for things like Social Security and Medicare, we should seek to reduce the cost of raising children, which can be done through things like lowering the cost of housing, providing more support for children’s health, and decreasing the cost of education. Of course, this is what many families are trying to do when they move to suburbs at the edge of the city. Ideally, policies to decrease the cost of raising children would be targeted at low-income and middle-income families, since there are many more of them than high-income families, for whom the cost of having children is probably less of a concern anyway.

Lastly, California is simply not scarce on land. We’re not Singapore or Hong Kong, not even by an order of magnitude. Santa Clara County has a population of 1.9 million, for a density of 1,400/sq mi. That makes it about 65% as dense as LA County, which includes enormous swaths of unpopulated land in the San Gabriel Mountains and Mojave Desert, yet still has over a million and a half single family homes. You can go 30 miles from Palo Alto and find vacant land and farms – well within the reaches of a Chicago-style extensive commuter rail, if that’s your fancy.

Like some parts of California, Singapore and Hong Kong have a lot of protected or unbuildable land close to the urban core. But unlike Singapore and Hong Kong, California is really, really big. For those who decide they must have a single-family home to raise a family, even if we upzone everything in the LA, you’ll always have the Antelope Valley.

Freeway Ramp Removal: The 2 at Fletcher

One of the undesirable features of some of LA’s older freeways is substandard spacing between successive interchanges. This creates congestion on the mainline, because there’s not enough weaving space for traffic coming onto the freeway to cross lanes with traffic getting off the freeway. Since LA’s arterial roads are often spaced a mile apart – about the ideal spacing for urban freeway interchanges – the additional ramps often connect to local streets that shouldn’t serve traffic accessing freeways. Think, for example, of the 101 between downtown and the Hollywood Split, the 10 between the 710 and the 605, the 110 between the 105 and downtown, or the 5 between the 710 and the 605.

This results in a situation where an urban improvement can coincidentally benefit a freeway, or where a freeway improvement can coincidentally benefit the city, depending on your point of view.

A local example that would be a great candidate for removal is the interchange of the 2 and Fletcher Drive, on the border between LA and Glendale. I suspect that this was a temporary south termination of the freeway built in 1958, extended across the river to intersect the 5 in 1962, but I can’t find confirmation.


This interchange provides very limited function for traffic circulation. Traffic exiting the 2 southbound could use the San Fernando off-ramp, located just 2,000’ upstream; traffic entering the 2 northbound could simply continue up Fletcher, turn right on San Fernando, and use the San Fernando on-ramp. This situation creates a dangerous weave on the 2 southbound, made worse by the fact that most of the traffic on the 2 south is trying to get to the right hand lanes, which lead to the 5 south. This interchange also takes up about 800’ of property fronting on the LA River.

Given the low utility, there’s no reason we can’t simply eliminate this interchange and redevelop the property. This is a great site, fronting on the river, and since there’s a large high voltage power line running on the river bank, the riverfront will remain green space. The rest of the site should be zoned R5, the city’s highest residential zone. A very rough layout is shown below.


The grey areas show roads; the orange creamsicle blobs are building sites. Note that we can reuse the underpass for the northbound onramp to provide an access to the opposite side of the freeway, where there’s some additional vacant land and a mini-storage facility that could be redeveloped too. At R5 density, these three sites could accommodate about 380 housing units, and of course, density bonuses could increase that further. That’s a pretty decent contribution towards the mayor’s housing goals.

Technical note: while this sketch is crude, the roadways were laid out using DraftSight, a free 2D CAD program. If you want to play around with horizontal construction like roads & railroads, but don’t have access to AutoCAD or MicroStation, I cannot recommend this program highly enough. It has a very AutoCAD-y feel and saves in dwg format, and you can reference in aerial images. Like the basic version of AutoCAD, it doesn’t do railroad spirals, but there are pretty easy workarounds for that, at least for rough planning. Unlike Google Earth & Sketch Up, this is easy to use for horizontal geometry (what, you don’t know the chord length or “bulge” – i.e. middle ordinate  – of the curve you’re trying to lay out?) and will satisfy curiosity towards the engineering feasibility of a preliminary design.

LACMTA Bus Ridership Update: July 2015

Another three months has passed, so it’s time for another LACMTA bus ridership update. As always, we start with the raw data. Highlighted cells represent the top 10 months for that route (since January 2009).


Here are the weekday, Saturday, and Sunday 12-month rolling averages.

bus-wkdy-12mo-201507 bus-Sat-12mo-201507 bus-Sun-12mo-201507

There’s not much new to say, so we’ll keep it short. Lines that have seen slight decreases continue to decrease; those that are steady seemed to keep holding. The Silver Line continues to grow slowly.

Here’s the percentage of trips on each arterial being served by the rapid route.


The share of riders served by the rapid routes continues to slowly rise on most corridors. This doesn’t necessarily mean increasing ridership on the rapid – it could be that both the rapid and local declined, but the rapid was more resilient. For example, here’s the split for Wilshire, where the Westside local (Route 20) has been fairly steady, the Rapid (Route 720) has seen a modest drop, and the heaviest drop has been on the east side local (Route 18).


That’s it for now; next up, Valley bus ridership, and then Metrolink. I’m going to keep these ridership posts short and go to a 6-month update cycle, in order to allow more time for more interesting posts.

Civil Engineering: Easiest to Understand, Hardest to Do Right

When I joke about why people should go into civil engineering, I usually say because it’s the easiest branch of engineering. And from a technical perspective, it’s true. You’ll never have to do a Laplace transform or find that equations with imaginary numbers are the easiest way to understand the system. Unless you’re deep into materials science, you won’t have to understand much at a molecular level. Unless you’re a structural engineer, you won’t have to do very much in the way of mechanics. You won’t have to worry about designing and building things with dimensions accurate to the thousandth of an inch. Civil engineering has pretty high tolerances. No one here ever got hosed by being off by 10 nanometers.

The math behind civil engineering is really not very hard. If you understand simple geometry and algebra, you already know most of what’s needed. Many things we do, such as drainage design, deal with chaotic systems like turbulent water flow, which are best explained by empirical relationships worked out long ago. Like anything, some people have a natural talent for seeing how a road, rail, drainage system, or other facility should be laid out. But once you’ve laid it out, the calculations to see if it works are much simpler, at least in my mind, than trying to analyze something like an integrated circuit or an eccentrically braced frame.

However, civil engineering is unique in that the public plays a huge role in the day to day operation of our systems. Consider, for example, the electrical system in your house or apartment: you are given one voltage, available at fixed locations, with the interface designed to make it almost impossible to do it wrong. Or take an airplane or a chemical plant: use of the thing is restricted to a small group of trained people. Or something like a computer or a television: an extremely limited set of functions is available to the typical user (someone who doesn’t take them apart).

On the other hand, consider a road: practically anyone can get a license, and when they get out on the road, can do damn well near anything they please; just about everyone, of any age, is potentially a pedestrian. Or take a sanitary sewer system: people can, and will, put all kinds of things into a sewer besides sewage. Or a grade crossing: it is almost impossible to physically keep people off a railroad crossing when a train is coming.

This means that a large part of civil engineering is trying to understand human behavior. A design might make perfect sense from a technical perspective, and no sense from a human perspective. There are many classic examples of designs that are nominally better in the technical sense but worse in practice. A case frequently found in the US is the missing crosswalk – an intersection where crosswalks are only provided on three of the four approaches. This design works technically in that it increases the auto capacity of the intersection by eliminating a conflicting or time-consuming pedestrian movement, but practically it puts a perilously large amount of faith in the pedestrian to execute their part of the system according to the plan.

Beg buttons, especially in high-volume pedestrian areas, are another design that, um, begs to be violated. No doubt you have jaywalked when encountering such intersections. Transit stations that only allow entry at one end of the platform when the other end would also serve desired pedestrian movements are another example. These are not good designs and the fact that the pedestrian movement is illegal doesn’t make the design any better. You wouldn’t design a kitchen stove that set the walls on fire if you turned it up too high and then depend on the user to not turn it up too high.

The extensive interface with the public is one of the most challenging aspects of civil engineering, and it can be tempting to try to ignore it. Ultimately, though, we need to do better at relating our designs to expected behavior, because that will truly determine the success of a project.