Category Archives: Innovation

When is advocating for a diversity of school models immoral?

In many cities I work in, education reform leaders bang the drum of needing more diversity in school models in their cities.

“Choice isn’t choice if all the schools are the same,” is a common statement.

I get where these folks are coming from, and I hope the breadth of effective school models continues to grow. There are still too many kids who don’t thrive in the schools that exits today, and I’m glad there are entrepreneurs who are developing new models, as well as investors and intermediaries that are supporting them.


At the same time, in many cities, two conditions are present:

1) Tens of thousands of children are attending failing schools.

2) Effective high expectations / high support (formerly called No Excuses) schools have waiting lists.

In other words, families are stuck in failing schools and existing high-performing schools could expand to help them.

Of course, just because these conditions exist doesn’t mean that it’s immoral to invest in new school model design.

But the “choice isn’t choice” refrain feels pretty privileged.

When your child is in a decent school, diversity of school model might be exactly what you’re looking for in a better option.

But when your child is in a failing school, you’re often just looking for an effective school who will nurture your child’s academic and personal growth – and if that school already exists in your city, you’re simply trying to get in.


To put a rough marker on there, if the aforementioned conditions exist in your city (many students in failing schools, many good schools with waiting lists), I think you should be devoting a good ~80% of your philanthropic funds on getting kids out of failing schools and expanding schools that are working.

If you’re allocations are reversed, and you’re spending the majority of your resources on new school models, I think you’re actions are not in the best interests of the children who are being harmed in failing schools.

If I had to argue against myself, I’d say that increasing the diversity of school models will increase the diversity of parents sending their children to public charter schools, which will strengthen the pro-charter political coalition. I’d also argue that new entrepreneurs and new models might pressure the incumbents to continue to adapt. Moreover, I’d argue that a real new school model breakthrough might prove to be more scalable than existing models, so investing in new models might help more kids sooner than scaling existing models. Lastly, I’d argue that all children, not just children in failing schools, deserve great school options.

I do agree with these counterarguments, but, for me, the near term weight of the moral good still sits with helping the children who are currently stuck in failing schools.

Hence my 80% / 20% calculation.

So yes, let’s keep on trying to develop new school models, but let’s make sure to check the privilege of making this argument with too much force, especially in cities where existing good schools have room to serve more kids who are stuck in terrible situations.

Personalized learning is a transformative idea without a transformative technology

Screen Shot 2017-07-11 at 4.59.53 PM.png

I just got back from vacation, which was a great time to read the The Three Body Problem science fiction trilogy, a wonderful series that revolves around the protoganist using first principles thinking to negotiate with an alien species.

Upon return, I read this Rand report on personalized learning, which was funded by the Gates foundation. The report covers a small set of schools in the early years of implementation, so best not to draw too firm of conclusions.

The report found:

  • Charters that adopted personalized learning strategies saw a +.1 effect in math and no statistically significant in reading.
  • District schools (very small N) saw no achievement gains.
  • Charter schools implemented personalized learning strategies with more operational fidelity.

Perhaps most interestingly, the authors noted:

In this theoretical conception, schools that are high implementers of PL [personalized learning]  approaches would look very different from more traditional schools. In practice, although there were some differences between the NGLC schools and the national sample, we found that schools in our study were implementing PL approaches to a varying degree, with none of the schools looking as radically different from traditional schools as theory might predict.

So in this sample, charters outperform traditional schools (thought by a lesser margin than urban charters as a whole outperform traditional schools); charters execute better; and the schools themselves don’t look radically different than traditional schools.

Hence the title of this post: personalized learning is a transformative idea without a transformative technology.

Without a technological breakthrough, the current personalized learning efforts will, at best, lead to modest improvements on the execution of common place ideas (using data to drive instruction, executing leveled small group instruction, investing children in goals, etc.). School will look the same and be a little more effective and pleasant for all involved.

This is fine and the world is in many ways built on modest improvements.

But for personalized learning to live up to its hype (as well as to its philanthropic investment), it will need a technological breakthrough.

Instructional platforms might be the first breakthrough, but even here I think the primary effects will be more around scaling great school models and content rather than deep personalization.

The crux of the issue is this: computers are simply not as good as humans in coaching students through instructional problems.

Your average person off the street remains a more effective grade school tutor than the most powerful computer in the world.

Until this changes, personalized learning will never realize its promise. The problem is one of technology, not practice.

What if unified enrollment platforms were 10x better?

An emerging group of cities – including Washington D.C., Newark, Camden, New Orleans, and Denver – have adopted unified enrollment systems. With these systems, families can enroll in schools across the city via an online application system.

This is a huge step forward. For too long, parents have not had enough information or access to the public schools in their cities.

However, the new enrollment systems are still in their infancy. The best version of these systems could radically improve public education. Unfortunately, we’re very far from this endgame.

I. Early Wins: Access, Equity, and Ranking

Access: With the best open enrollment systems, families who can’t afford a house in a fancy neighborhood can now finally transparently apply to a school in a more wealthy neighborhood.

As a result of increase in access, a recent study in Washington D.C. found that the new enrollment regime would likely reduce segregation over time:

Screen Shot 2017-03-17 at 9.23.22 AM

Equity: In most cities, traditional and charter schools play a ton of games with enrollment. Traditional magnet schools use opaque entry requirements. Bad apple charters don’t take in kids with special needs. There is no equity.

With online enrollment platforms, these problems go away, as schools are no longer in control of their enrollment.

Quite simply: the algorithm is fairer than the enrollment clerk.

Ranking: These new enrollment systems also allow parents to rank their top schools. This is extremely important.

First, a family’s high desire to enroll their child in a school can now  be translated into an increased chance that they actually get into this school.

Previously, high desire meant little unless you were connected, wealthy, or dogged.

Second, ranking allows  parents to publicly signal to government which schools are most and least in demand (which will ideally affect opening, expansion, and closure decisions). It also signals to school operators what attributes make a school in high demand.

By analyzing ranking preferences, researchers in New Orleans were able to correlate school characteristics with parent preference:

Screen Shot 2017-03-17 at 9.44.39 AMRanking transforms family desire into actionable information.

II. Unified Enrollment Systems are Mediocre Platforms

In preparation for writing this blog, I spend an hour on unified enrollment system websites. It was not a great experience.

Here is the school finder homepage from Washington D.C. – I couldn’t even find a way to filter schools by academic performance!

Screen Shot 2017-03-17 at 9.50.14 AM

Here is the search function for Newark’s enrollment system – you have to download a pdf!

Screen Shot 2017-03-17 at 9.55.36 AM

By comparison, here’s the search page from Zillow:

Screen Shot 2017-03-17 at 9.53.10 AM.png

On Zillow, you can easily search for homes based on the most useful search criteria. Yes, Zillow is surely better capitalized than your average enrollment system, but even with modest funds a city should be able to do better than a downloadable pdf.

III. Moving From Equity and Ranking to Matching and Prediction

More sophisticated uniform enrollment could offer two extraordinary improvements: they could better match families with schools, and they could better predict how any given student would do at a school.

Matching: Right now families mostly use enrollment systems for ranking: they know the schools they want and they use enrollment systems to express this desire.

What is not really happening (as far as I can tell) is sophisticated algorithms actually helping families match with schools.

For example, on Camden’s enrollment site (where you can thankfully filter by academic performance!), I found three schools that all met the “on track” performance criteria, and pulled up the comparison page:

Screen Shot 2017-03-17 at 10.11.14 AM

This does very little to help me choose between these schools. My preference variables are limited to very broad categories such as “art classes” and “after school programs.”

After playing around on these websites, I get very little indication that that the platform knows me or the schools very well. Compare this to Netflix, Amazon, or dating websites (previous life) – platforms where I feel like the algorithms know me better than myself.

Unified enrollment systems need to more deeply understand children and schools in order to make better matches.

Prediction: Right now, government accountability systems are a basket case of poor design (generally don’t weight growth enough), brutal politics (what politician wants to tell a bunch of communities they only have “F” schools?), and awful transparency (good luck trying to navigate your average state department of education website).

Most importantly, government accountability systems evaluate schools rather than make predictions.

As a parent, it’s one thing to tell you that a school is a “C+” – it’s another thing to give you a prediction of what will happen to your child if she attends the school.

With current date, we could probably gather basic information on your child’s age, gender, current academic performance, personality type, etc.,  and make a reasonably accurate prediction that if she attends school X she will have a Y% chance of graduating from high schools and a Z% chance of earning a post-secondary degree.

Good enrollment systems, over time, should become better and better predictive agents, and, perhaps, can end up augmenting (displacing?) government accountability systems.

IV. Root Causes and Potential Solutions 

I don’t yet have strong beliefs about the root causes of why these enrollment products aren’t getting better faster. But here’s some guesses:

Non-profits > government operated: Most of the enrollment systems are run by governments, which are not good at running tech products and have bad incentives around giving parents accurate information about schools. Non-profits would likely be better operationally and have better incentives, and avoid the privacy concerns associated with for-profits.

Lack of scale: Matching and predication can better with bigger data sets, and if all these systems are structured as isolated city based data silos, the algorithms will be dumber than they should be.

Weak Customer Demand -> Bad Economics: SchoolMint, from what I understand, is the most successful player in the market. For reasons I don’t underhand, this company has not developed a better product. Perhaps it’s because their government customers don’t actually want it. Or perhaps the economics don’t work (which might suggest philanthropy is needed).

If the above is true, a national non-profit should be backed to scale to enough size to create smart algorithms, and it should be financially structured in a manner that gets it out of the perverse incentives of being beholden to government or individual schools rather than families.

A philanthropic foundation with a great tech backbone could be well situated to support this endeavor.

V. Expectations

Better matching and prediction would probably not make the average student’s education experience 10x better, just as dating websites don’t inevitably lead to great marriages.

But I do think better matching and predication could increase the probabilities that millions of families could find a better fit for their children.

At scale, that’s a better world.

School supply or regulatory change – what should the Trump administration incentivize?


Andy Smarick has a good short paper out where he gives advice to the new Trump administration.

He proposes a $250 million program that would fund “high-quality, high demand, highly accountable programs.”

In short, Andy proposes a federal fun to increase the supply of things like course choice, education savings accounts providers, voucher-based private schools – as well as the creation of innovative regulatory regimes that can monitor the performance of these entities.

I really like Andy’s suggestion that the federal government continue to invest in the creation of great providers for the following reasons:

Launching new schools and courses costs money, and given that most providers are non-profit organizations, they need funding in advance of receiving regular per-pupil revenue.

The federal government lacks knowledge of what local communities need; by funding entrepreneurs, the feds are backing those who can only succeed if they meet local demand, thereby reducing the risk that funds will work against local needs.

Innovation has positive externalities: it’s well known that, in the private sector, entrepreneurs rarely capture the full value of their innovations. The same holds true in the public sector: great non-profits do not capture all of the social good they create; rather, it leaks into the full  system over time (i.e., Teach Like a Champion).

I’m a little more skeptical that the government should incentivize the creation of new accountability policies. Whatever you think of Common Core or teacher evaluations, it’s unclear that the federal governments involvement has been a long-term net benefit for the sustainability or effectiveness of these policies.

Yes, Andy’s suggestion is much less heavy-handed. He’s only arguing that the federal government should fund accountability pilots.

But I worry that this is simply reinforcing a bad habit.

My current thinking is that the federal government should focus solely on research, information transparency, supply, and civil rights.

Of these four, supply is currently the most under-appreciated and underfunded, and Andy is right to call for an increase in funding here.

Should Ed-Tech Platforms Empower or Restrict?

I’ve previously written on being bullish about the potential of ed-tech platforms.

Currently, both Summit Public Schools and Alt Schools are leading the way on developing platforms that may eventually be used by thousands of schools across the country.

Many people are drawn to ed-tech platforms because they can: (1) support teachers to curate innovative lessons and execute more personal coaching; and (2) allow children to learn at their own pace and explore their intellectual interests.

In short, ed-tech platforms are about empowerment.

But it is unclear to me that empowerment will be the only way that ed-tech platforms improve education.

I think they might also improve education by restricting educators and students.

I’m still trying to work through this, but see below for a graphic representation:

Screen Shot 2016-07-20 at 8.39.49 PM

The goal of many (thought not all) personalized and ed-tech enthusiasts is to move from wherever they are to the top right corner.

This vision has much to be said for it, and under the right conditions it very well may work.

But there is also another option – one based more on restriction than empowerment. A couple of great educators have been pushing me to think about this path as well.

The argument for restriction goes something like this:

  1. The No Excuses charter movement has learned a lot about what it takes to increase the learning of students who are multiple grade levels behind.
  2. It will be very difficult to scale No Excuses charter schools due to human capital, operational, and political constraints.
  3. Professional development has proved generally ineffective in spreading the practices of No Excuses charters to mediocre charter and traditional schools.
  4. A tech platform that utilized software that mimics the instructional practices of No Excuses charter schools – and then frees up teachers to do scripted small group and individual tutoring – could be a way to scale the core components of the No Excuses model while bypassing traditional human capital, operational, and political constraints.

Under this scenario, the goal is to move from the bottom-middle row (I do think No Excuses charters are empowering students more than before) to the top-middle row (with more scripted curriculum and teaching structure preventing this model from being ed-tech progressive).

In this model, the tech platform is really a backend way to scale a high-performing whole school model, in that the platform would dictate curriculum, assessments, pacing, and staffing.

Ideally, this packaged model would only take up 3-4 hours a day, and there could still be plenty of time for true project based instruction, extracurriculars, etc.

In summary: perhaps there is a (mostly) best way to teach basic reading and math, and, perhaps, a tech platform can scale this (mostly) best way.

And maybe the “big data” from such a platform could further evolve the (mostly) best way.

I’m not really sure. All feedback welcome.

Who Will Education Platforms Liberate?


I’ve been living in San Francisco for a few months now.

During this time I’ve had the chance to talk with some great educational entrepreneurs who are making different platform bets.

A platform is a plug-and-play business model that allows multiple participants (producers and consumers… who may be one in the same) to connect, interact, and create value.

Education platforms are varied.

Some are content neutral: numerous programs can plug in and users can access in any way they want.

Some deliver more standardized content: fully baked competency curriculum, tasks, and assessments – with  more heavy curation of user generated content.

What I’m most curious about is this: who will education platforms liberate?

Platforms could liberate students. They might be better able escape mediocre curriculum, weak assessments, and substandard teachers and get better instruction, psychological development, and career guidance through platforms.

Platforms could liberate teachers. They might be able to better escape terrible district mandates and simply close their doors, plug into the platform with their students, and execute a far better instructional model.

Platforms could liberate school founders. The barriers to entrepreneurship could significantly decrease if a new school is plugged into a platform that does a lot of the heavy lifting in terms of technological, operational, and academic infrastructure.

Of course, platforms could end up liberating them all: students, teachers, and school founders could equally benefit.

On the other hand, platforms might also not deliver and simply liberate investor of their money and educators of their patience.

Who Will Innovate in the Education Sector? The 1%? The 2.5%? Everyone?

A new report – “Dissatisfied but Optimistic” – just came out. You should give it a scan.

The paper applies a theory of change for innovation to the education sector. Here it is:

Screen Shot 2015-09-04 at 7.01.26 PM

You’ll see that the model is predicated on a small number of innovators creating new models that then get scaled.

Here’s how they predict the adoption curve will look:

Screen Shot 2015-09-04 at 7.03.02 PM

This theory of change, perhaps*, conflicts with another theory of change currently being applied in the reform field.

This theory of change goes like this: if we create incubators, accelerators, design shops, innovation fellowships, etc. in many cities across the country, we’ll launch a new wave of educational innovation in schools across the nation.

So which will it be? Will innovation be driven by the rare disruptors or by the empowered educator force? Or somewhere in-between?

I’m not an expert an innovation, but it does raise an interesting resource allocation for the reform movement: what percentage of funds should go towards capitalizing the coasts versus empowering the fly over states? Should we be concentrating resources in the top CMO (existing or emerging) management teams in the country or should we fund citywide school design efforts where small teams of principles and teacher leaders are supported to innovate in their own schools? Should we be investing in the top 1% of entrepreneurs or should we fund teacher innovation fellowships?

You get my point. I don’t have strong opinions on this. Others might.

But here’s my initial guess: breakthrough whole school model innovations will come from the top 1-2% of all-star management teams. Modest improvements in existing systems may be achieved by empowering and supporting innovation at the teacher / school leader level. But I’m not convinced the latter work (design thinking training, innovation fellowships, etc.) will lead to major innovative breakthroughs, though it may end up being useful for creating the stage for early adoption.

What do you think?

What am I getting right or wrong?

*You could make the argument that the “teacher design training / innovation fellowship” model could be used to identify the 2.5% of disruptors. But that’s not my read on it. Rather, I think it’s more about empowering teachers to innovate in their own settings, not become the next great educational entrepreneur.

Elon Musk and Other Rich Parents are Finally Beginning to Subsidize Education Innovation Rather than Simply Purchase Status


I previously wrote about how wealthy families negatively impact the rest of the world when they spend a lot of money on education status rather than education innovation.

When this occurs, wealthy families purchase zero-sum goods (status) instead of subsidizing goods with positive externalities (innovative education models).

Yesterday, I wrote about how scaling charters and vouchers to the suburbs could be one way to increase both quality-options and innovation to a diverse array of kids.

Today, I read this: Elon Musk created a small school for the children of employees of Space-X.

With Alt-School and Musk School, we’re beginning to see wealthy families investing in new education models.

If these models work, they will overtime benefit all children.

Will these models work? Who knows. But one way to begin to find out is to experiment on some rich children.

Yes, you can get grumpy and say that these private schools (right now) only serve wealthy students, but in the same breath please do say how you support vouchers programs that are designed to give all students access to these schools.

Big picture, I’d much rather have wealthy people building innovative private schools than sending their children to stagnant schools of status.

Thiel, Cowen, Taleb, Options, Candler, Tiny Schools, Regulation

micro school

I just read this transcript from Tyler Cowen’s interview of Peter Thiel.

I found myself not being able to fully evaluate many of the claims that were made. I don’t know if this is because Cowen and Thiel are simply operating from a much deeper knowledge based than myself, or because they were just being vague and non-linear. It is likely some combination of the two.

Anyways, I did agree with this from Thiel:

If you are starting a computer software company, that costs maybe $100,000, to get a new drug through the FDA, maybe on the order of a billion dollars or so. If the FDA were regulating video game technologies, and you had to do a double-blind study to make sure that the video games weren’t addictive, damaging to your brain, etc.

In short, there are some sectors, predominantly health and education, where over-regulation hinders innovation.


I’m currently reading Black Swan and AntiFragile by Nicolas Taleb. (HT Ethan Fletcher)

Taleb has so many strong opinions, some of which seem clearly false, that it makes hard to know which of his ideas to trust. But many of his ideas do seem insightful.

I was particularly drawn to his emphasis on options as being a model for how we should think about innovation. Basically, an option minimizes downside while maximizing upside, in that, for a price, you get the option to buy at a certain price without having to commit to buy at a certain price.

His point is that with innovation you similarly want to minimize downside risks while maximizing upside opportunity.

The VC sector basically works on this model: make a bunch of (relatively) small bets, all of which have very high upsides, and hope that one of them works.

In this sense, options are a way of incorporating intellectual humility into bets about the future.


Recently, Matt Candler has been advocating for Tiny Schools.

One way to think about Tiny Schools is that they are just that: schools with not a lot of kids. An entrepreneur opens a Tiny School, educates some kids, and hopefully the school succeeds.

Another way to think about Tiny Schools is as options: an entrepreneur starts a schools with 15-20 kids and tries to get the model right. If the school doesn’t work, 15-20 kids get a subpar education experience for a few years. Not really the end of the world, though of course not ideal. If the school does work, then the entrepreneur can attempt to scale the model, either by opening more Tiny Schools or by converting the Tiny School into a Regular School.

In short, launching Tiny Schools minimizes downside risks (less kids, no need to buy a large facility) and maximizes upside potential (the schools that do well can then be scaled).


What regulatory structure might allow for Tiny Schools to thrive?

One model is just to keep Tiny Schools private. This is fine, but requiring tuition inevitably limits who will be served.

Another model is to use public funding, but to make the regulatory model very thin.

I discussed this with Adam Hawf and landed at the following:

  • Market Size: A state would limit the amount of Tiny Schools that received public funds in any given year: perhaps 100 Tiny Schools with no more than 50 students each in any given year. This equates to roughly 5,000 students /  $50 million a year, assuming $10K in public funding per student. This limits the downside for taxpayers.
  • Entry: The application process would be a slimmed down version of a charter application. The entrepreneur would need to have a budget, brief description of the model, and a basic opening plan. 30 pages at most.
  • On-going Regulation: On-going regulations would be based on homeschooling regulations. Basic reporting on students served and finances, but very little oversight.
  • Back End Accountability: There would be no back end accountability. Zero testing requirements or anything of the sort. Rather, the license to operate a Tiny School would be a one time, four year license. Once the license expired, the operator would either need to apply to become a charter school or a private school. From there, the school would enter what ever accountability system the state used for the chosen sector.


If even ten states adopted this model, at any time we’d have a 1,000 different Tiny School pilots going on around the country.

The downside would be very small.

The upside could be incredibly high.

Why So Little Innovation in Education?

A Young Lady's Illustrated Primer

At this point, it’s nearly obligatory for a conference speaker or panelist to say either:

(1) “Classrooms look the same exact same as they did one hundred years ago.”


(2) “We need to transform our outdated factory model of schooling.”

Rarely does the speaker then give a plausible explanation for why classrooms look the same now as they used to do a hundred years ago, or why we haven’t thrown off the awful shackles of the factory model of education.

If we’re going to increase education innovation, we should probably try and understand why there is so little of it.

So let’s dig in.

Some possible explanations for the lack of innovation could be:

1. Regulation is hampering innovation.

Education in this country is heavily regulated, especially at the K12 level. It’s possible that this regulation is blocking potential innovation. Regulator inhibitors likely include: monopolized school operations, unified standards and annual assessments, and lack of family choice.

2. Lack of radically innovative talent.

The education industry might be failing to attract radical innovators into the field, perhaps because of some combination of culture, structure, or prestige. The true innovators might be drawn to other fields.

3. Lack of a profit motive.

A profit motive can spur innovation, especially when it comes to rapid marginal improvements to existing products (think Moore’s law). The lack of profit motive in education could be a factor in reduced innovation.

4. People don’t really care about learning.

Despite all the fuss about achievement, perhaps people mostly just care about things like status, peer groups, and safety– so the producers of education respond to this demand. Parents may be both value non-academic qualities and be very risk adverse, each of which could hamper innovation.

5. There’s not much low-hanging fruit.

Perhaps schools still look the same because the way we are doing it is the best way to educate children for a reasonable price. Limits on human capital and technology might make it (currently) difficult to radically improve them model.

 6. There’s actually a lot of innovation happening.

One might argue that innovation is occurring – it’s just less visible. Advances in data-driven instruction might be an example of a “silent” innovation.

So which is it? 

I imagine most people would point to the first four causes (regulation, talent, no profit motive, status seeking) if they were pushed to explain why innovation is lagging in education.

And I think each of these causes does play a role.

But there are also pretty big holes in each of these explanations:

If regulation is hampering innovation, then why aren’t private schools innovating more?

If a lack of talent is hampering innovation, how to square this with the fact that TFA out recruits most industries when it comes to Ivy League talent?

If lack of profit motive is hampering innovation, then why don’t we seem more innovation in textbooks, computer programs, and school operations – all of which have significant profit motives?

If people don’t really care about learning, then how come the recent New Orleans study on parent choice demonstrated that academic performance was the amongst the highest rated determinants of school selection?

There’s not much low-hanging fruit?

I’m not sure, but perhaps.

I will say this: I think there’s less low-hanging fruit than most people think there is.

That being said, I do think all of the above factors are real, and, to some extent, are inhibiting innovation. I just don’t think any of them are slam-dunk answers.

Perhaps if we continue to make headway in areas such as deregulation and talent, then we’ll see much more innovation.

But there’s a chance we might not.

It’s possible that the real innovation breakthroughs won’t occur until artificial intelligence is much more sophisticated, or the cost of labor is dramatically reduced.

My guess is that both of these circumstances will occur in the next fifty years.

*This post is based on a conversation I had with a charter leader in Memphis. As per usual, talking to awesome people continues to be a primary source for my thinking.