The sewers I swim in

Boston sewer image from Liquid Assets

I've seen lots of arguments about why reducing the deficit right now would bring crisis to the economy. Most of them are very textbook Keynesian arguments arguing that at times of excess capacity, reducing deficit spending would just add headwinds to an already struggling economy. The other argument is that the US should take advantage of exceptionally low borrowing rates to invest in rapidly aging infrastructure and put Americans back to work using a sort of New Deal 2.0 scheme.

The first argument is a bird's eye solution to a ground-level problem. Yes, government spending would goose GDP, but is that spending creating wealth? Where is that "stimulus" going? Our goal, after all, is not to maximize GDP, but to maximize wealth. GDP is just a poor objective measure for a deeply subjective phenomenon and gaming our own framework won't help anyone, regardless of what numbers the BLS, BEA and FRB release over the upcoming months. And let's not forget that Washington has a very poor track record as an allocator of capital. I'm simply not comfortable leaving these decision up to the people that decided to try to reflate the bubble by pulling-forward demand, subsidizing toy arrows and foreign liquor and build useless airports. Just sayin'.

But does this mean we should address the crisis with full-throttle austerity? Not quite. As it was eloquently pointed out last Summer on interfluidity, austerity is stupid and deficits are dangerous. We can't make generalizations about debt, deficits or balanced budgets. Deficits and debt are neither good nor bad on their own. Leveraging up for wealth-creating projects is good, borrowing to throw money away shoveling sand from one pile to the other not so much. Washington is focusing on abstract goals like "putting real Americans to work." And one can't blame them because that's what people want, jobs. But "jobs" isn't something you can simply create from thin air, you can't just throw money at this problem and expect to fix it. "Jobs bills" and "improving America" are nebulous ideas, subject to interpretation without any objective way to measure success or failure, which is probably what Washington wants.

"Well, fine, but what do you suggest then?" you may be asking yourself. I just want to say one word to you. Just one word. Sewage. We've spent the better part of the last 10,000 years trying to secure sources of clean water and get rid of waste. Humanity has developed modern plumbing and sanitary sewers. We survived the Great Stink of 1858. We've battled epidemics of water-borne disease, droughts and floods.  I feel comfortable in making the broad statement that clean water is good and shitty water is bad. Therefore, one could expect that making something good out of something bad would be a positive thing, an improvement, a wealth-creating action. If you disagree, feel free to stop reading now.

All of which brings me back to our original topic, the deficit. We have swaths of unemployed persons and slack capacity in all aspects of construction, record-low financing rates, and an economy that uses fresh potable water faster than it replenishes it (including aquifer sources). Wouldn't it be great if we could put excess capacity to work creating an infrastructure that helps us achieve sustainability and conserve one of our most vital resources while financing it all at record-low rates? Well, we can, and it's called sewage treatment. It's the effective, efficient and inexpensive process of cleaning water.

The Deer Island Treatment Plant on the Boston Harbor provides primary and secondary treatment for the waste and storm water of the greater Boston area. It serves 43 communities, 2.5 million people and hundreds of thousands of businesses and it cost $3.8 billion to build. The entire MWRA had $176M of sewer-related operating expenses in FY2010 (pg 50). That works out to $70.40 per-person per-year. That figure includes not only the plant, but the entire sewer system as well as treatment of storm water and one of the most advanced plants in the country. DITP not only discharges water that is cleaner that the water it is being discharged into, but it efficiently decomposes organic waste using anaerobic digestion, reducing the volume of the sludge by 90% and using the resulting methane gas to help heat/power the plant. The dried, pelleted result of the digestion process is sold as Bay State Fertilizer (the heat naturally created by the digestion and drying process kills the harmful pathogens). So, for $70.40 per-person, per year, the MWRA cleans, on average, 360 million gallons of waste-water per-day and turns the organic water contained in it into energy and high-quality fertilizer, saving the city millions of dollars in sludge transportation and disposal fees, all while keeping the harbor clean. And while the $3.8 billion cost of construction may sound like a lot, consider that the plant had an initial 30-year expected life, meaning buying the plant on credit and amortizing it over 30 years (using the current 30y tsy rate of 4.12% as a proxy) would cost a only $7.36 per-person, per month. To put it all in perspective, including both amortization costs and operating costs, the cost per-thousand gallons of water treatment comes out to $1.48, or about the price of a medium-sized water bottle in a convenience store.

That's deficit spending I can get behind.

UPDATE-1: fixed an arithmetic oops and added reference to plant cost.
UPDATE-2: It was pointed out to me that the actual number of users serviced is 2.5 million, not 2 million. All numbers adjusted to reflect this. Link to source added as well.

Wikipedia, Khan Academy and The Price of Knowledge in the Digital Age

Being twenty-six years old and having been heavily-involved in technology since my early teen, I've had a chance to witness a lot of big things from the beginning. I remember when Audioscrobbler (the thing that drives Last.fm) was still a grad-school project.  I remember when OpenID was just a post on Brad Fitz' LiveJournal. I remember when Wikipedia got it's 1000th article. My first contribution as a registered user dates to September 2002. Back then Wikipedia was a sort of wild-west and I didn't even know what "peer-review" meant. I've been programming for the better part of the last 12 years and and remember making contributions to the software that runs Wikipedia before it even had a name. I remember the site being down for hours and hours--sometimes even days--when the servers were overloaded or a hard drive failed.

Anyway, one of my favorite memories is showing my father Wikipedia and him laughing. He attempted to explain the concepts of authorship and authority, why he was skeptical about the quality of the site, and how the concept of it all was a radically different from the status-quo--naturally, I had no interest. As time went on, he increasingly became a believer too. We were seeing this massive revolution happen before our eyes, in the shadows of the .com crash, when nobody gave a flying fuck about tech anymore. I didn't realize what it all meant then, but I remember my father talking to his friends about it and heated debate happening; they were all genuinely interested and excited about what it meant. In the true, original spirit of the internet, Wikipedia was democratizing access to knowledge, leveled the playing field like few things before, and it was all capital "F" Free. The price to access it was $0.00 and the value of it was increasing exponentially as the network of information nodes became increasingly interconnected and of increasingly better quality. The most relevant comparison in my mind is probably the Gutenberg press.

Recently, I had a feeling of deja-vu as I signed-on to the Khan Academy. I first learned about the Khan Academy a little over a year ago. I was having a little bit of trouble remembering some mathematics concept and someone suggested I look for it on Youtube, as there was this "Sal dude" posting instructional videos. I searched, found, learned, and didn't really give it much thought again. Since then, Sal Khan was gotten a ton of press, and so recently, idle from my lack of employment, I logged on to the Khan Academy. I was totally shocked by what I found. Not only is the library of lessons huge but I finally used the problem-generating component, and I was totally floored by what I found. Sal created a "Knowledge map" (accessible once you sign-in with a Google account) wherein as you complete certain lessons and the set of problems satisfactorily, new areas are suggested.

The system is basically a tree of nodes that correspond to a specific lesson, each of which has an instructional video. Each node has zero or more parents, with the root node being "Addition 1". As nodes are completed, new ones are recommended, organically building upon the previous ones. Completing "Addition 1," recommends "Addition 2" and "Subtraction 1". As nodes are completed,  the student can continue to explore each branch of the tree independently. Likewise, someone who is interested in learning, for example, linear algebra, could look at the map and follow it back until he or she found the first familiar subject and then beginning with the next lesson.

The videos--mostly math-related at present time--are both focused and engaging. Khan is a gifted teacher, able to distill lessons to their core and present everything you need to know about one thing in a few minutes.  With 20-30 minutes a day, one could easily follow the knowledge map and become proficient in most any mathematics subject in a couple of weeks or months. Any motivated learner has the ability to learn whatever they want, at whatever speed they want, at no cost at all. For parents who lack the knowledge to coach their children or are unable to afford tutors; adults that need additional education but lack the economic means or time to do it at a traditional venue; and students wishing to place higher in college math-placement exams in order to save themselves an unnecessary and expensive introductory or remedial course, the implications are huge. Khan has singlehandedly, in a remarkably short time, changed the landscape for mathematics education.

To test it all out, I decided to "attend" the Khan Academy. I started with lesson 1, "Addition 1." and worked myself up to calculus over a couple of days. I didn't watch the videos for the simpler subjects like addition and subtraction, but I did start skimming through once I got to Algebra II and started really watching in the later parts of Trigonometry. I was amazed of how fast time went, and how rewarding it was to complete the subject examinations--you must get 10 consecutive questions right in order to advance. Both the videos and problem-sets were generally completed before my attention span was exhausted, which means 10-15 min. As I completed problem sets and worked my way up the map, I started feeling like I was opening new levels on a video game or something, it was really strange.

The process is not limited to individuals working on their own. Pupils are able to enter a "coach ID" and be linked to a "class". The interactive problem sets generate data that the coach can use to identify students that are stuck on a certain type of problem or are working at below/above average speeds. Because the students can work independently of one another, no one student is held-back or left-behind. Someone having trouble with, for example, variable substitution can simply re-watch the short lecture, rewinding or fast-forwarding as needed to focus on the points he or she doesn't grasp.

I understand that this is not the first experiment of it's kind, I'm familiar with Open Course Ware and I've used iTunes U, but this is definitely different. The casual approach, accessible language, narrow focus, instant availability, lack of requirements and flexible structure all combine to create something orders of magnitude more accessible to the casual user. Not everybody knows how to use a podcast, but everyone knows how to work Youtube. Just click "play," and look at your screen, it's that simple. Have a question? Under each video is a list of previously-asked questions and responses, and if yours is not on there, you can add it instantly and someone is likely to respond promptly. I'm not saying this approach will work for every subject, but I have no doubts it could work for at least undergrad-level physics, chemistry and finance. As someone who barely passed high-school chemistry, something like this would have saved me a lot of angst as I tried to cover a month's worth of skipped classes the Thursday night before the test.

With the recent press the Khan Academy has been receiving, including public accolades from Bill Gates and sizable donations, the project has been able to pay Sal a salary and will be able to fund it's continuing existence. As momentum builds, it is not unreasonable to expect the number of contributing teachers to increase--some volunteers are already helping translate and close-caption the videos. Because the subjects covered don't change, there is no reason why these videos couldn't be used for generations to come--although, admittedly, some of the earlier ones could use a quality upgrade. The value produced by the Khan Academy is accumulative and increases with each additional topic and translation--I can't even imagine what a textbook company would be willing to pay for it--yet the price to access it, like Wikipedia, is zero. Of course, the free ability of content online is not enough, you have to give people access to the internet first, but with the ubiquity of mobile-phone service, rapidly-falling prices of computers and initiatives like OLPC, it isn't hard to imagine a world ten years from now where 80% of school-age children have access to the internet, even if it is from a shared device. When I first understood what Wikipedia was about, nine years ago, all I could think about was "Wow, this is going to change everything." Mark my words, this will change everything, too.