Where Can You Find MOOCs From the Top Universities?

I’ve compiled a list of the top 25 universities according to US News & World Report, and which MOOC Provider you can find their courses on, using the data from curricu.me.




Princeton Coursera
Harvard EdX
Yale Coursera & Yale Open Learning
Columbia Coursera
Stanford Coursera
University fo Chicago Coursera
Duke Coursera
Upenn Coursera
California Institute of Technology Coursera
Dartmouth College iTunesU
Johns Hopkins Coursera
Northwestern University Coursera
Brown University Coursera
Washington University in St Louis Semester Online*
Cornell EdX
Vanderbilt Coursera
Rice Coursera & EdX
University of Notre Dame Semester Online
Emory University Coursera & Semester Online
Georgetown University EdX
University of California – Berkeley EdX
Carnegie Mellon Open Learning Initiative
University of California – Los Angeles UCLAExtension
University of Southern California n/a

*Semester Online is a paid, for-credit service with an application process, so it does not qualify as a traditional MOOC

Some quick observations:

–       University of Southern California is the only holdout on this list.

–       It really speaks to the power of Coursera and EdX. There are some young providers like Canvas or FutureLearn who will need to find a niche or have a really great value proposition in order to be successful in this market.

–       It also speaks to the power of MOOCs. 24 of the top 25 universities offer MOOCs or MOOC-like courses.

Amazon Prime Air Drone

The Economics Behind Amazon’s Drone Delivery Service

“Amazon will be disrupted one day…I don’t worry about it because I know it’s inevitable.” These were the words of Jeff Bezos as he talked about the eventual death of Amazon. He said it calmly, like it didn’t keep him up at night. But with the pace and the grandeur of innovations coming out of Amazon, you might think he is working with the frenzy of a man trying to avoid death.

Amazon’s latest idea is an autonomous flying drone that delivers packages within as little as one half hour. In other words, Amazon becomes more instant than a drive to and from the mall.  The drones look like rather normal octocopter drones fitted with a box-grabbing mechanism on the bottom and a covered area on top which holds the motor and presumably some complex computing hardware.

A quick look at Twitter and some other articles shows that after people get over the Jetsons factor a common question is, “well how much does it cost?”



I’d like to take a look at the cost of drone delivery versus human delivery using simple back-of-the-napkin calculations.

The cost of human delivery

To determine the cost of human delivery, we need to compute the cost per package. Let’s determine the daily cost of a UPS delivery person and divide that by the number of deliveries he makes per day, or:

Cost of human delivery


Rate of driver: According to Glass Door, a UPS driver makes about $26-$28/hour. Let’s round it down to $25. Then let’s factor in gas, tolls and parking tickets and bump that back up to $30/hour.

# of Packages Delivered: According to a UPS driver forum, a typical driver delivers up to 250 pieces per day. Seems like they also work about 10 hours per day. I won’t calculate overtime.

So then the cost of moving a package from UPS center to front door is about $1.20 per package:

Cost of human delivery is 1.2


This does not include the cost of getting the package to the UPS center or the cost of whatever sorting UPS has to do, or any other overhead. So I presume it is on the extremely low side of things.

The cost of drone delivery

To figure out the cost of drone delivery, we also need to determine the cost of delivery per package. Let’s determine the lifetime cost of the copter, and divide that number by how many deliveries a copter can make in a lifetime, or

Cost per package - Amazon Drone


Cost of drone: An octocopter on 3D Robotics with autopilot and GPS navigation costs $1,000.  Amazon’s drones must also include the grabbing mechanism and presumably a more intelligent computer that can avoid other drones or hitting people as they land. Let’s say that Amazon drones cost $3,000, and they last 5 years. After maintenance and electricity costs let’s bump that number to $4,000. I won’t include the cost of whatever infrastructure it takes to manage this fleet of drones.

# of Packages Delivered: The maximum range of the copter, according to Amazon, is 10 miles. We’ll set the average trip at 5 miles one-way, so 10 miles round trip. And we still need to calculate how many trips they can do in their lifetime.

There is an interesting site which measures top speeds achieved on multi-copter drones. It says that 10% of pilots achieved speeds of over 50 mph. This seems very fast for autonomous drones zipping around above urban areas, so let’s say that they travel 25 mph on average. Most helicopters on the site seem to reach these speeds, and this is also about how fast they would need to travel in order to complete a 10-mile delivery within 30 minutes.

At these speeds, with a 5 mile average trip and time given for loading/unloading, a drone can make 2 trips per hour.

Drones are not confined by human working hours or working days, so I presume they would deliver during all times that a person might normally order a package. Let’s say 14 hours per day and 6 days  per week (a day off for maintenance). So in 5 years a drone can make 43,680 trips

Screen Shot 2013-12-03 at 6.37.59 PM


Our final calculation becomes the cost of the drone divided by the number of packages delivered:

Cost of Amazon Drone per trip is 9 cents

The estimate cost of drone delivery is 9 cents per delivery. Human delivery was $1.20 per delivery.

There could be errors in the numbers above. I put in no costs for additional infrastructure. I didn’t consider the cost of skilled employees who need to maintain and manage the fleet of drones. I don’t know if there are things like airspace fees, etc. But what I do know is that the lifetime cost of the drone would have to be greater than $50,000 to even be on par with the cost of human delivery. And the difference between drones and humans is that the cost of drones goes down over time…

The Conclusion

There are still a lot of questions here. Foremost in my mind is how many customers it will reach. With a 10-mile range and 14 Amazon Fulfillment centers, the drones can reach 0.14% of the land in the United States. Given that these fulfillment centers are not in urban areas, I presume that means that very few customers are in range. It sounds like Amazon will still need to leverage UPS or USPS to make this work, at least as “drone headquarters”.

Nonetheless, just back-of-the-napkin numbers make a strong case that this is a shrewd cost-saving move by Amazon. It is also a stark example of the lengths that organizations will go to in order to minimize high labor costs.  The lesson here is that if we are not constantly improving our skills, then our job will be outsourced or automated, because it just makes sense in a competitive free market. The consolation prize though, is that at least you can get those new sunglasses you ordered within 30 minutes, without even leaving your computer.  



The Nerd’s World Cup: Does Behavioral Economics Disprove Rational Economics?

The USA just beat Mexico to qualify for the World Cup.

From a rational, economic perspective, the USA should qualify for every sporting event on the planet. The USA has more people, more resources, and more free time to invest in sports. When you take a look at the major medal winners in the Olympics, more medals mostly correlates to these factors.

So while the fact that we are the bottom-feeders of soccer doesn’t hurt my American pride, it does pique my interest. Why doesn’t the rational model work?

The answer is that the rational model does work, but the number of factors that I put into my model while walking from the TV to the bathroom doesn’t even come close to telling the whole story. There are many more factors: The United States lacks a history of soccer, we lack a culture that cares about the sport, Sportscenter doesn’t air it enough, Mexico’s star player had a tight hamstring (maybe). If we could measure these and the 10,000,000 other factors that go into it, you could literally chart on a graph how many wins a soccer team earns.

In an economic decision the same simplification errors apply. We say that people will pay $3 for a bottle of Gatorade because it gives them $1 of hydration, $1 of flavor, and $1 of brand appeal. Allegedly, any purchase can be rationally broken down into the elements of value that the purchaser receives. Obviously, this doesn’t explain the many abnormalities in rational economics, like why someone will pay $6 for the same Gatorade. These abnormalities are explained by the science of Behavioral Economics. A behavioral economist will explain that one person buying that same Gatorade for $6 can be attributed to the fact that the person has a higher price tolerance, due to gradually being accustomed to higher purchases. Or maybe even because someone told him Gatorade is worth $10 and he thought getting it for $6 was a great deal. That behavioral economist might explain the USA’s overall soccer futility with the theory that emotional benefit of winning a soccer match is not as great the emotional benefit of winning a football match.

And if they did this, I would agree with this economist. But then that economist might continue his theory, and try to say that behavioral economics disproves rational economics. Because an emotional factor affects team USA’s performance or the price we pay for a Gatorade, we therefore can’t create rational models. With this I disagree. Behavioral economics does a good job of identifying cost/benefit factors which are hard to quantify. But just because factors are hard to quantify does not mean they are not actually measurable. We could, with enough data, measure how the number of seconds that Sportscenter airs soccer coverage affects soccer enrollment among children. And then (theoretically) do this for 10,000,003 other factors. When we arrive at the end of this excruciating process we have, once again, a rational economic model for the number of wins a soccer team earns.

Behavioral economics is a stop-gap. It is a reasonable way of way of making shortcuts in measurement when you simply lack the tools to measure further, like saying the world is flat because it looks flat. But it still a part of economics. And, like team USA, it has a ways to go before it reaches its goal.

Theory: Greater Social Touch leads to Higher Reviews

I’ve recently been on ODesk, AirBnB and Skillshare, and noticed that these websites are full of glowing reviews. But I’ve also seen some downright nasty reviews on sites like Yelp or Amazon. Clearly, there is a correlation between higher reviews and how much personal interaction you’ve had with a service provider. But how strong is that relationship and what, specifically, are the drivers? Is it conversation? Is it seeing someone else’s picture or sharing yours? Is it not really driven by social kindness, but rather just a reflection that personalized service is generally better?
It’d be an interesting experiment. Let me know if someone has done it. If not, I may try it in the future. 

Chromecast is a Competitor’s Nightmare

Sorry for the missive that is about to follow. I have a thing for Google, and I’m writing this in a fit if adulation after reading about the new Chromecast. If I had kids, I’d imagine its akin to watching a great milestone in a child’s life, like a first step or a first word. I’m just a blubbering state of emotion.

Very short introduction: The Chromepass is a $35 USB stick that transmits video to your television, same as Apple Airplay.

Here is why I like it so much: it solves a very prevalent problem for an incredibly reasonable price. This is the how to build products, and the confirmation is in the fact that is has sold out of Amazon and Best Buy in just a few days.

Google could have gone about this in another way.  Google could have tried to undercut Apple just enough to capture some market, and then try to maintain that market through artificial barriers.
Seeing that Apple TV is $100, Google could have priced their product at $90, and then created artificial barriers to cross-sell other Google products, like say that content could only be streamed from approved apps and video on the Chrome browser. This would have been a perfectly intelligent short-term option, and one that would probably be seen as generous by most people, including the author. Apple TV would cut their price, Google would follow, Apple would release a further-reduce Apple TV lite, Roku would build their own, until finally five or six PC-to-TV streamers would compete and they’d all cost $30.

Now we get into fuzzy numbers. This process might have taken 5 or 6 years. During that time, Google would have been making higher margins on their product. They may have recouped their investment during that time. But in the end they would be one of six other competitors in the same space, fighting for the same slim margins. By starting out at such a surprising price point, they capture huge market share very quickly, and they force competitors to either put out products that aren’t ready or put out a product too late. In a market with this much potential, early mistakes could be costly. So long as Google delivers a good product experience, which they usually do, this looks like a truly shrewd move.

Georgia Tech takes a step towards the future, and answers a lot of questions

Georgia Tech announced last month that it will offer a Master’s Degree in Computer Science, delivered as a MOOC (Massive Open Online Course) via Udacity.com. It is the most high profile institution to date to offer a full online curriculum for a degree. And it will cost under $7,000 per year, dropping to around $1,000 per year in the future.

Often, decisions that others say are “complicated” or “too long term” are determined when a first-mover says “f@*! it” and goes for it. Such is the case with Georgia Tech. This move answers a lot of the questions that have surrounded MOOCs from their inception. Here were some of my pressing ones below:

Will universities offer their content as MOOCs?
This question has been answered pretty convincingly. The rise of websites like Udacity, Coursera, Udemy offering content from Harvard, Stanford, MIT is solid proof of a popular idea. Now the business model needs to be able to sustain itself.

Why would a university offer credit through a MOOC course?
The answer here is simple. Businesses that can offer the right product, for the right price, to the right customers succeed. A university degree is a product, just like a bar of soap. There is an opportunity for Georgia Tech to make a highly demanded product available to a wider group of people, and likely make more money from the efficiences of online education. This is what the internet does for many products, and I see little difference with college degrees.

If a university offers a MOOC, why would students pay for the brick-and-mortar experience?
Brick-and-mortar will always have a place. There is still value in a true college experience, with equally intelligent and inquisitive people, and close proximity to professors and laboratories. But in the future this value will be reserved for those who are willing to pay for it.

What about exclusivity?
Georgia Tech has said that they aim to admit “all applying students who satisfy the basic admissions prerequisites and qualifications”. This is pretty radical, because I imagine that most people measure the value of an education with its exclusivity. But exclusivity is actually a fairly weak and artificial measure of value. It means there is a lack of supply. For a university, supply is limited by the space within their walls, the number of desks they have, faculty, etc. Businesses should be trying to increase their supply without sacrificing the objective value of their product, and only leaning on scarcity as a marketing tactic when that is no longer possible. Georgia Tech made the right move in not imposing an artificial exclusivity barrier.

How will they pay for it?
Once again, the first-mover didn’t overthink it in this case. In the first year, the course will be subsidized in the same way that most things on the web are paid for: by sponsors and advertising. After that, GA Tech is relying on the assumption that people are willing to spend $7,000 (or less) on a year of college credits. When an equivalent brick and mortar degree will cost closer to $50,000, I think this is a safe bet. 

Why Computer Science?
Computer Science lends itself well to online education. The answers are black and white. Tests and quizzes are easily graded for large groups of people. Computer science students are web-savvy. The content and the participants work well with online collaboration and forums. Computer Science is an easy choice for a pilot program. If successful, though, the model can still be applied to other degrees as well.

Now for some more academic questions…

Can a 12-year olds earn masters degrees?
Well not yet. GATech’s FAQ page states that all participants need to go through an application process and submit things like previous degrees. But looking towards the future, it is an interesting question. I see little reason why a 12-year old who graduates from a course like this wouldn’t be able to receive a degree.

Will it be successful?
A document exists that shows GATech’s expected revenue and expenses for the first three years. 
By some back of the napkin calculations:

  • GA Tech expects 208 students in year 1, 141 in year 2 and 2,907 in year 3. 
  • AT&T’s one-time $2,000,000 investment is marginal. By year two, the program will be self-sustainable and by year three costs will reach almost $14.5 million. Revenue will be $19 million, without any corporate sponsorship. 
  • According to the document, Georgia Tech will receive 60% of the revenue, but pay only 17% of the costs. The other 83% are paid by Udacity. If that is true, then Udacity is running at a $4.2 million deficit in year three. Either this is bad information, or Udacity is making a bold move to step out ahead of competitors like Coursera. 

Our economies most lucrative profession…is free

Two articles ran today, one in the New York Times, one in the Wall Street Journal. The New York Times piece was about the effect that student debt has on the economy, where debt-laden graduates postpone consumption and major purchases, creating a drag on the economy. (Eerily similar to my piece about the same topic, I’m checking my Google Analytics now). The second piece from the Wall Street journal was an op-ed by an entrepreneur who says that his business, and the economy in general, can’t find qualified computer scientists fast enough to fill the job openings.
These two pieces are coincidentally recent, but the opportunity they describe is old news. The most common complaint I’ve heard from entrepreneurs is that they can’t find–no, can’t afford–good programming talent. When I was a business student at college, non-students would beg to get access to the tech groups at universities to pillage the talent. Google and Facebook are this countries most lavish employers for a reason. According to a very quick look on indeed.com, computer scientists now make more money than financial analysts, which is where most of my finance friends are now.

The most surprising fact here is that computer science is very nearly a free degree. The web is full of free programming training, free user forums, open-source code. And proving one’s computer-programming proficiency is considerably simpler than proving one’s proficiency with more vague skills like marketing. This should make that brand value of a degree from a university less valuable. From an economic perspective any cost-benefit analysis where the cost is nearly free and the benefit is very great should be over-saturated with people.
So why the gap between the number of lucrative openings and the number of applicants, which Kirk McDonald from the Journal article puts at 3 to 1? First, I simply think that the expansion of computer-related needs is just that rapid. Sometimes I think we probably can barely produce enough graduates in general to keep up! (In fact, if we consider Americans, we can’t. We now borrow from every other country.) Second, I think it is the devilish economic problem of not-quite-free information. There is still the concept of doctor-lawyer-banker professions being the best. Doctors are still doing ok, but bankers are looking more plebian and lawyers can’t find jobs. More importantly, the free path to success as a computer scientist is more uncertain than the expensive college degree path. 
Dispelling that uncertainty and creating freer information is doable, especially in a field like computer science. Two things need to happen:
First, the free education that is available needs to be aggregated and sorted. A college education exists online, but it is probably on 12 different websites. 
Second, avenues need to be provided where competent programmers can prove their proficiency. A combination of a online certificate of achievement for the courses and some kind of proctored exam or interview to prove real knowledge, could tell an employer most of what they need to know. 
If these two things can happen, we could end up in a situation where a large portion of our workforce is placed into lucrative, high value jobs with little personal debt. 
If you want to be this person and build this for others at the same time, e-mail, write, call me collect. 

Cooper Union and the Grandeur of Free

Cooper Union, a small engineering and arts school in the East Village, has made the announcement that they will begin charging tuition to students who have less need for financial assistance. The tuition will go from $0 for all students, to a maximum of $20,000 for an unspecified percentage of students.

This sounds like an inevitable decision for a small university based in the heart of New York. Studios in the area that rented for $400/m twenty years ago are now renting for $2,500. Property taxes are following the same rapid trend. Universities are in a kind of “arms race” to build the best facilities in order to attract the greatest talent. Presuming they’ve got some kind of health plan for their employees, they’ve picked a deadly trifecta of rising costs

The Deadly Trifecta

It is also a catastrophic decision for the appeal of an otherwise small and frugal school. In 2011, Cooper Union had a lower acceptance rate than Yale, Princeton and Brown. They pride themselves on being one of the most diverse of the smaller schools, both culturally and economically. With only 1,000 students and a few old buildings, this is an anomaly that can be greatly attributed to the appeal of “free”.

We are attracted to free in a much greater sense than we are attracted to “very cheap”. Would the line outside Ben & Jerry’s be the same length if they advertised “$.01 Cone Day”? No, it would be much shorter. Cooper Union has done their version of “50% off Cone Day”. A $3 ice cream cone is still not a great deal.

In a battle of market forces versus historic character, I am of the mind that the market wins 95% of the time. The old opera house will eventually come down so that the high rise can be built. So it is in this case. But I don’t think that Cooper Union has considered the magnitude of how this changes their university. In this case, we may be looking at a market decision that will eventually lead to an institution’s demise.

Prediction: Would not be surprised to see them reverse this decision. Why don’t they just put ads on all the chalkboards and pre-load every student’s phone with Facebook Home?

Is Crowdsourcing Irrational?

I’ve been spending a lot of time lately trying to put together a website, and in doing so I’ve thought a lot about the effects of crowdsourcing. When I have a question about how to do something (which is every time I build a new piece of the site), I can always find the answer online in places like Stackoverflow or the WordPress forums. Usually I can find the answer with one Google Search, and very often I find that the entire functionality is pre-built with something like a public Git or a WordPress plug-in. This content is usually high quality, often being verified by other members of the community, and it is always free.
Crowdsourcing is a fascinating thing for an economist, because it is such a large group of seemingly irrational people giving away their assets for free. Is there an economic cost-benefit equation behind crowd-sourcing, or does it break the laws of classic economics in a huge way? Here are three potential explanations, in no particular order:

  1. Crowdsourcing creates a greater eco-system for all involved. Think of “a rising tide lifts all boats” or the incredible amounts of money governments spend trying to recreate a Silicon Valley. A person who shares his work with others will benefit from the improvements of the entire community more than he will suffer from the effects of new competition.
  2. Crowdsourcers are just egotistical fools, so infatuated with getting their intelligence in a public forum that they will disservice themselves to do it. 
  3. Everyone out there is trying to promote themselves and realize long term benefits, much like a blogger who writes for free might do until he can sell his content for fees or ad space.
I would have a hard time believing that any one of the above reasons have given rise to the huge tidal wave of crowdsourcing. I’m either missing reasons, or true, unselfish altruism is alive and well on the forums of Stackoverflow. 

A Back-of-the-Napkin Case for Skipping School – Part 2

The rising cost of college has been well documented and debated over the last decade or so. Usually, these articles end with the obvious, and important question “when will it stop being worth it?” At some point, the cost of attaining an education will outweigh the benefit of receiving an education. I’d like to outline a holistic cost-benefit analysis, but using all back-of-the-napkin calculations.
Let’s start with some numbers. I will presume that a typical college education costs $150,000 after all aid, and that this debt takes 15 years to pay off. (This post is two-part, and part one explains how I got these numbers). I will also ask the web what the lifetime value of a 4-year college education is over a high-school education, and the Wall Street Journal states it is between $300,000 and $1,000,000. Let’s say $500,000, but we could easily redo this experiment with another number.

1. Real Costs and Real Benefits 
So far, we have a straightforward model. The cost of a degree is $150,000 and the benefit is $500,000. Net benefit is $350,000

2. Opportunity cost of earnings while in college
Continuing to be straightforward, we know that while serving 4 years at university, Facebook tells us that our non-university friends are making “straight cash” and buying way better cars than we have. The opportunity cost of those for years of earnings is, let’s say, $30,000 x  4 = $120,000. We’ll round down to $100,000.

3. Opportunity cost of Risk-Taking Ability
Remember that we’ve decided in part 1 that our graduate will be paying for his college education for around 15 years. This means that from the age of 22 – 37, our subject’s risk-taking ability is curtailed by his obligation to his lender. Let’s look at it a few ways:

  • The lifetime cost of a child is estimated at just under $400,000. Our student’s college loans are like having half a child at the age of 22.
  • Taking a simple average, our subject pays $10,000/year in college loans. If he instead invested $10,000 per year at a 5% return, he would end up with $225,000 at the end of 15 years. If he added no more principal after 15 years, he would still have close to $500,000 at the end of 30 years. 
  • According to Business Insider, the odds of startup success are 20%. The average IPO valuation is $250 million. I’ll resist the temptation to do an expected value on that (No I won’t! It’s $50,000,000.) But let’s be realistic and say that your ho-hum lawn care business can be sold for a $1,000,000. Expected value is $200,000. 
Despite these silly examples, debt servicing is the most damaging cost of college that gets sometimes overlooked. It creates a generation of non-innovators. And it increases in a non-linear way as the monetary cost of college increases. For a student paying $20,000 per year at 5% interest, it will take 3 years to pay off $50,000 but 41 years to pay off $350,000.

How we value this cost can be a judgment call. But I will use our third example from above and call it $200,000. For larger debt loads or a subject who has a higher tolerance for risk, I think this number could be much, much higher.

4. The WTHC Cost
Finally, the Who-The-Heck-Cares cost. A rational economic model states that we will work for our entire careers in order to realize as little as a $1 advantage over our non-college brethren. If the average lifetime income of a non-college educated person is $1,000,000, then a rational person should go to college if the average lifetime income after college costs is $1,000,001. As if in the coffin of our now dead college-educated person, he is wearing his alumni ring, holding that $1 and smiling smugly.
But less rational (real) people must ask themselves what the personal, non-monetary costs of giving up your 18-22 years to a university are, and then the cost of structuring your life around paying for those 4 years. Especially with the availability of substitutes which are available now, with free MOOCs or local community courses available for a fraction of the price, and social and networking opportunities like MeetUp.com allowing non-college attendees to be socially “plugged in”. 
Two factors, the rising costs of college and the greater availability of substitutes, have led us to a point where a traditional college education is on the cusp of not being worth it. And with the diverse number of factors that go into it, I can guarantee that many people are already overpaying. The student who does not qualify for financial aid, goes to an expensive school for a traditionally low earning potential career (art history?), and has an appetite for risk is almost definitely overpaying.
I would like to find the brave 18 year old who decides that a traditional education is not worth the rising costs, and this person will build their own curriculum and be better off than his traditionally educated peers.