Pur Autre Vie

I'm not wrong, I'm just an asshole

Monday, April 29, 2019

Citi Bike and Topography

As an exercise for the reader, see if you can identify which way the hill slopes in Crown Heights/Prospect Lefferts Gardens:

(click to expand)

I bet you could make a pretty good topographical map using nothing but Citi Bike data. Think how much easier it would have been for the colonists if they could have used Citi Bike data instead of tedious surveying!

Citi Bike Patterns: Tides and Rivers

It's a little after 7 p.m. now and the situation in lower Manhattan has changed a bit:

(click to expand)

As you can see, on net after the evening commute, Chinatown and the Lower East Side have gained bikes while the financial district has largely lost them (it's a bit mixed as you can see). Tribeca remains crowded with bikes. I assume this reflects commuting patterns, with Wall Street workers heading home and with Tribeca attracting an evening crowd for food and entertainment.

This is really the best case for the argument I made in my previous post for expanding the docks. Where stations are alternately full and empty, larger stations allow more people to use the bikes without adding much extra work for system employees. The analogy here is to a tide—the bikes come in, the bikes go out, all according to a diurnal cycle that naturally repeats itself. I'm not saying Citi Bike employees (and Citi Bike Angels!) never have to rebalance the bicycles, but in the ordinary course they just go back and forth with regularity. All you have to do to increase capacity is build more docks and add more bikes.

Stations that are perpetually full or perpetually empty are a harder case. In contrast to the tidal flow of lower Manhattan, the analogy here is to a river flowing from one station (or group of stations) to another. For instance in Park Slope most of the stations toward the bottom of the slope are in perpetual surplus, while the stations at the top of the slope (by the park) are in perpetual deficit. In effect there is a river of bicycles constantly going down the slope, presumably because people like to coast downhill but don't generally care to pedal uphill. Here's what it looks like (the slope is generally uphill as you go east):

(click to expand)

This is what Park Slope always looks like. I knew before I went to the map that I would find a suitable screenshot to illustrate my point. Individual stations vary, but the pattern of green stations at the bottom of the slope and yellow/red at the top is near-constant. (Occasionally the neighborhood gets inundated and everything goes green, but this never lasts long. By the way, that's one of the crappiest times to use the bicycles because there's nowhere to dock them. You end up going way out of your way and/or paying the fee for going over the time limit.)

If Citi Bike were to expand the stations in Park Slope, more bikes could fit in the lower part of the neighborhood, but within a day or two they would be full again and the ones at the top of the slope would be empty. It takes constant work to keep the stations in balance. It's not clear that increasing the number of bikes or docks would help much. The key problem is moving the bicycles each day. (With large enough stations maybe you could ensure that stations never run out of bicycles, but it would still take a lot of effort keeping them balanced.)

I don't have much of a point here, except to note that if Citi Bike were to expand its stations, it should expand them in "tide" areas more than in "river" areas. Station size is clearly the limiting factor in lower Manhattan, whereas in neighborhoods like Park Slope the key thing is simply to move the bikes uphill, against the current.

Expand Citi Bike Now

I continue to free ride on the boundless optimism of tech investors as I cash in on the Citi Bike Angel program. Those gullible capitalists will soon be buying me a coveted white Citi Bike key—which I will get completely gratis when I hit 500 angel points! This is obviously unsustainable, but I intend to take advantage to the hilt while the party lasts. By the time they sober up, the coveted white Citi Bike key will be in my pocket, and I'm not giving it back.

Anyway as a result of my lucrative Citi Bike Angel activities, I've noticed that quite frequently docking stations are completely full or completely empty. This makes them unusable for people who want to drop bikes off or pick bikes up, respectively. Here's an example of what I mean, from late this morning (this is a very typical pattern for a day like today):

(click to expand)

The stations that are entirely green are completely full, the stations that are red are completely empty. Yellow means that a station has 1-3 bikes. (3 bikes is represented by either yellow or green depending on the station, I think.) The pattern here is obvious: Chinatown and the Lower East Side are net exporters of bikes on a weekday morning like this, while the financial district and Tribeca are big net importers. This almost certainly reflects commuting patterns (although bear in mind, the bikes from the LES didn't necessarily go to the financial district and vice versa). No idea what's going on with DUMBO, I've never paid much attention to the stations there.

(By the way, those numbers you see on the stations reflect Angel points ripe for the taking. If you take from a station with a white number on a black background, and drop off at a station with a black number on a white background, you get the sum of the points indicated. You can also pick up or drop off at a neutral station for a smaller point gain. This is what I mean when I say that my lifestyle is financed by the naïve optimism of the investing class—look at all those opportunities to earn 2 or even 4 points in a single trip! And as a reminder, it takes only 500 points to go from the ordinary, undistinguished blue Citi Bike key to the coveted white Citi Bike key.)

Anyway the important thing to note here is that there is almost certainly a lot of unsatisfied demand for both bikes and docks. It's unlikely that everyone on the LES who wanted a bike this morning got one, and it's unlikely that everyone who wanted to dock a bike in the financial district was able to do so. People are having to walk quite a distance to find the nearest bike (or walk quite a distance from the dock to work), and presumably a lot of people who would like to use Citi Bike to commute are unable to do so and are using other transit modes. (Also, anecdotally, when I'm chasing Angel points I often see someone immediately take a bike I just dropped off, or immediately use a dock I just freed up. So in my observation there is definitely unmet demand for those things.)

This is a failure! It's always going to be the case that bikes are relatively numerous in the financial district on weekdays after the commute, and likewise for docks in places like the LES, but if the stations were doubled or tripled in size you would observe a lot fewer stations that were completely full or completely empty. Or if you didn't... it would mean (in rough terms) that 3 times as many people were commuting by bike, which would be great in itself.

The only cost (aside from the docks and bikes themselves, which I assume are not expensive) would be some parking spaces. But bikes are vastly more efficient than cars in terms of the parking space required. The space taken by a single car could easily provide parking space for 8-10 bikes. So trading away those parking spaces should be a big net improvement for mobility in the city, even putting aside the vastly better affordability of Citi Bike. (You could also expand parking for private bicycles, but I think Citi Bike is a superior option for most people and I would favor it over racks for private bikes. But with sufficient political will you could have plenty of room for both.)

Biking is healthy, it's affordable, and it causes far less air pollution and congestion than car travel. The city would reap large benefits by allowing Citi Bike to double or triple the size of stations in high traffic areas.

Friday, April 26, 2019

Why Didn't the Seabird Fly Over the Harbor?

There is nothing so frustrating to me, being a vegetarian, as unwrapping a bagel that is meant to be my late breakfast, having eaten nothing so far, to discover that liberal amounts of bacon have been placed on the cheese, making the bagel suitable only for the dumpster.

What is the current woke take on giving such a bagel to a homeless person? Would this be hopelessly condescending? Would it wrongly treat homelessness as an issue for charity instead of correctly analyzing it as a political issue? I don't know. Anyway I missed my opportunity because while I was on the subway two homeless people came by but at that time I didn't know about the bacon. (I probably would have risked my woke credentials by offering the bagel.)

Wednesday, April 24, 2019

How to Engineer a Product

A long time ago I read a Malcolm Gladwell piece about Ron Popeil, the television marketer. It made a point about capitalism that I've thought about from time to time ever since.

Let's start by imagining a company designing a product. Let's say it operates in a competitive sector, so it has to cater to consumer tastes. (Even monopolists have some incentive to cater to consumer tastes, but it will be clearer for a competitive or at least quasi-competitive market.) As a starting point, we can imagine that the company designs its product to maximize its usefulness to customers, but limited by some kind of budget constraint. So for instance, an engineer might determine that making the walls of the product out of stainless steel makes it durable and easy to clean, and this may be justified by the cost of the steel. So the firm incorporates this change to its product design. And more generally, this is the sort of change the firm is looking for. Design innovations should generally look like improvements noticeable to the consumer (even though the consumer may not understand the engineering considerations involved).

But with a little more thought it becomes clear that there is another category of design change that we should expect to see from time to time: one that doesn't improve the product except maybe by lowering its price. An engineer might determine that two materials are equally functional in the final product, but one of them makes the product much easier to assemble in the factory. Or there might be a custom part used in the product that can be replaced by a standard (and therefore much cheaper) version, and this might be worth it even if it makes the product slightly less functional. The cost savings may or may not be passed along to the consumer.

So now we've got a potential source of compromise in terms of product quality. You can imagine products getting worse over time as engineers develop more and more cost-saving techniques. Also, material costs change over time, and so do labor costs, and so do shipping costs, and so on. So optimizing the product design becomes a matter of chasing a moving target, and something like the deregulation of trucking could end up influencing the design of consumer products. At this point engineering has become quite a bit more difficult than simply making the best possible product subject to a budget constraint.

We can introduce another complication, consisting of a difference between what is best for the consumer and what the consumer perceives to be best. Or, similarly, consumers may have aesthetic tastes that, if catered to, compromise the functionality of the product. To make it even more complicated, perhaps only a subset of consumers have these beliefs or preferences, while the rest are agnostic (or perhaps have correct beliefs about product functionality). In that case you may simply write off the deluded and/or idiosyncratic consumers, but conceivably you may decide it's more profitable to manufacture an inferior product (from the viewpoint of some or potentially all consumers) in order to broaden the market for your product.

I hope it's obvious by now that it is potentially a very tricky thing to design a product. Of course you have the option to manufacture multiple versions of the product, in order to satisfy different segments of the market, but this comes with its own costs.

None of this is my point, though. As Gladwell points out in his piece, one of Popeil's big insights was that product design and marketing are not distinct. You don't design the best product possible and then figure out how to market it. You design the product to make it easy to sell.

This is subtly different from my earlier point about consumers' aesthetic preferences. What I am talking about now is not just giving the product an attractive appearance, but giving it attributes that specifically enhance its marketability as opposed to its functionality, cost of manufacture, or conformity to consumer tastes/beliefs.

The example from the Gladwell piece is the large, slanted glass door on Popeil's rotisserie, which allows the television audience get a good view of the food cooking inside. This may or may not be a desirable feature in a rotisserie, but it (apparently) makes it easy to persuade television audiences to order the product, and that makes it good design from the business's perspective.

What is a consumer to make of this? I don't know. I think you should be at least a little uneasy about a product that has been designed to push your buttons. And I think it's reasonable to worry about the compromises that have been made in functionality simply to sell you the product. (Remember, we're specifically talking about enhancements to the product's marketability, not to its appearance once purchased. Though I suppose serendipitously these could sometimes be coextensive.)

[Edited to add: A great example of this phenomenon is the shiny red apple (Red Delicious?) that became ubiquitous in the United States. It looks like the prototypical "apple," shiny and red, with few blemishes, but it tastes like cardboard. Of course this product was not designed, but it was subjected to selective pressure, which amounts to the sane thing.]

Now consider that this is how ideas work. What makes for a successful idea? Well it might contain truth and therefore perform some useful service to the people who believe it. (I guess this is pragmatism.) Or it might flatter people's prejudices, or it might simply be beautiful to them.

But ideas can also simply, by design or by luck, have features that make them attractive to adopt without any other usefulness. Once this occurred to me I began to see nonsense everywhere, nonsense with "viral" attributes. We all believe others to be vulnerable to manipulation but we believe ourselves to be impervious to it. It doesn't seem possible that we can all be correct in this belief.

Now I should note that our prejudices and our limitations of intelligence and patience are probably enough to ensure that nonsense will flourish. You don't need this virality concept, this idea of building an idea to make it maximally infectious to the human mind. But it doesn't help, and anyway for whatever reason the world is full to the brim with nonsense, and this makes life much worse for most people, but I guess better for the Ron Popeils of the world.

Friday, April 12, 2019

Expand Social Security Now

I have thought a bit about universal basic income (UBI), and I don't have any strong views. It is a policy with pros and cons, some of which are difficult to assess (for instance, as a matter of politics, UBI appears to be an inefficient way to "spend" whatever political capital exists for redistribution—but maybe this would change if it were actually enacted and people felt its benefits). It's very hard to come to a firm conclusion.

But one observation I'd make is that the case for making Social Security (A) universal and (B) considerably more generous seems very strong. Basically this is UBI for retired people, and it seems like a great idea to me. (Part of its genius is that you don't have to call it UBI! It's already got a name, one with a lot of goodwill attached to it.)

A major failing of capitalism is that people's welfare in retirement is dependent on saving a lot and managing the money prudently, while avoiding market crashes, fraud, and other threats. It also depends on people retaining their capacity to manage their investments well into senility, which is highly unrealistic. (As a side note, I expect that in the coming years we will see massive amounts of fraud against elderly people. It's just such a rich vein to mine and the victims are so ill-equipped to protect themselves.)

Of course anyone who wants to save for retirement should be allowed to do so. But we shouldn't have a society where you have to be a savvy investor to avoid penury. Social Security has been very successful at reducing extreme poverty, but it is not generous enough for most people, and it has big coverage gaps (the amount you get is a function of how much income you earn during your working years, which means some people aren't even eligible).

By the way, Social Security has a really nice feature, which is that you can't outlive it the way you can outlive your savings. It will be there for you for your whole life no matter what. This sort of product is available from the private sector (it's called an annuity), but as you would expect the market has all kinds of bullshit products and ways to get fleeced. Social Security is rock solid by comparison, and it's absurdly cheap to administer.

Also my understanding (this is not legal advice!) is that Social Security benefits can't be encumbered and can't be garnished by creditors (except maybe the IRS?) or bankruptcy trustees. That's a really nice feature. In essence, most seniors would be entirely sheltered from financial calamity (or to put it a different way, even if financial calamity strikes, the consequences would be far less severe than they presently are).

Apart from the fiscal expense, I don't really see any downsides to expanding Social Security. (As for fiscal expense, payroll taxes currently apply to the first $100,000 or so that an individual earns. This cap could be lifted to generate a lot more revenue. I don't know if that would be sufficient, however.) I suppose the politics could be a bit tricky, especially if undocumented immigrants were covered, but this seems like a minor problem at worst. For the most part I expect the politics would be very good. There might also be knock-on effects—for instance, the less the typical voter cares about the stock market, the easier it should be to impose stricter regulation, higher capital gains taxes, and the like. (I admit that is somewhat speculative, and of course lots of middle class people would still invest.) Also I have to imagine that the infusion of money into areas where the population is shrinking/aging would help some of the worst-off areas of the country.