A capital humanist: My best fiction reads of 2012

The hardest thing about writing is seeing clearly enough that you can get to the heart of an idea, a person in a sentence or two.  All of the authors listed here do this.  Their books are wonderful observatories of the human experience: some draw it more beautifully, some draw it more sharply, some more absurdly.  But every exaggeration reflects back some fun house mirror of the truth.

My favorite fiction reads of 2012:

• A Visit from the Goon Squad, Jennifer Egan.  Time is running out for faded rocker and record exec, Bennie Salazar.  As Bennie puts it,  “I’m done. I’m old, I’m sad – that’s on a good day. I want out of this mess. But I don’t want to fade away, I want to flame away – I want my death to be an attraction, a spectacle, a mystery. A work of art.”  Every chapter is a different window into Bennie’s life, from a different person, at a different time.  In the hands of a lesser writer, this might feel trinket-y.  But with Egan’s command of language, it’s just a treat to get to meet so many different people and see so many different times.

• The Invisible Circus, Jennifer Egan.  When 1960s San Francisco wasn’t enough of a kick any more for hippie Faith, she moved to Europe where she died shortly thereafter.  Her younger sister Phoebe feels like she’s missed out: missed out on the ‘60s, missed out on Faith.  Eventually, Phoebe takes off for Europe herself.  There she meets her sister’s ex-boyfriend, Wolf, living a rather pedestrian life as a translator.  Trying to understand the gap between the wild child lifestyle Faith and Wolf led together and the life Wolf now leads, Egan reflects,  “And sitting there, sea drifting in around them, Wolf had understood for the first time what kind of life he wanted to live with Faith. Maybe they wouldn’t rise up into the sky the way he’d thought, maybe the real thing was doing what his parents had done, pay the rent, read the paper, hell, maybe that was the dare. To live–day in, day out. Just live.”

• Look at Me, Jennifer Egan.  In this 2001 novel, Egan presages the rise of reality television, terrorism, and social media.  But like the other Egan books, this one explores what identity is and what it means when our identity changes.  As our protagonist says, “I would lie of course. I lied a lot and with good reason: to protect the truth—safeguard it like wearing fake gems to keep the real ones from getting stolen or cheapened by overuse. I guarded what truths I possessed because information was not a thing—it was colorless odorless shapeless and therefore indestructible. There was no way to retrieve or void it no way to halt its proliferation. Telling someone a secret was like storing plutonium inside a sandwich bag: the information would inevitably outlive the friendship or love or trust in which you’d placed it. And then you would have given it away.”

• The Adults, Alison Espach. The adults, consumed with their own wealthy suburban Connecticut lives, have more or less left Laura to raise herself.  The adults behavior is appalling, Laura’s own decision-making is questionable, but her tone of voice is more observational than excoriating.   Later, as an adult, she reflects, “Childrens’ lives are always beginning and adults’ lives are always ending.  Or is the opposite?  Your childhood is always ending and your adult self is always beginning.  You are always learning to say goodbye to whoever you were at the dinner table with.”

• Gone Girl, Gillian Flynn.  Razor-sharp writing with a killer plot.  Amy disappears on her fifth wedding anniversary; husband Nick is suspected. Amy writes of their relationship, “I am a thornbush, bristling from the overattention of my parents, and he is a man of a million fatherly stab wounds, and my thorns fit perfectly into them.”  But was it Nick?

• The Likeness, Tana French.  Daniel says to Cassie-as-Lexi, “Sacrifice is not an option, or an anachronism; it’s a fact of life.  We all cut off our own limbs to burn on some altar.  The crucial thing is to choose an altar that’s worth it and a limb you can accept losing.  To go consenting to sacrifice.”  Daniel and his housemates have chosen a different sacrifice than most.  But in the midst of it, the real Lexi has been killed.  This is a smart whodunit, exploring identity and critiquing modern life as its pages turn themselves.

• The Art of Fielding, Chad Harbach. A bildungsroman for the modern era, this one weaves together the stories of five college students, shifting points of view with each chapter.   The most fully rendered is Mike Schwartz, captain and unofficial coach of the Westish College baseball team: “Schwartz knew that people loved to suffer, as long as the suffering made sense.  Everybody suffered.  The key was to choose the form of your suffering.  Most people couldn’t do this alone; they needed a coach.  A good coach made you suffer in a way that suited you.  A bad coach made everyone suffer in the same way, and so was more like a torturer.”

• State of Wonder, Ann Patchett.  In the prologue to the Best American Short Stories of 2012, Geraldine Brooks writes, “There’s nothing wrong with writing stories set in bedrooms, classrooms, kitchens.  These are the places where we spend large slabs of our lives.  But the air becomes stale there.”  Patchett takes us out of these rooms and into the Amazon.  Against this canopy, a pharmaceutical company executive disappears, a researcher’s answers don’t make sense, and our protagonist seeks the truth.  Fingers crossed that after all the hard work Patchett does to set the scene in this book and the number of loose ends she leaves behind, a sequel is on its way.

• Super Sad True Love Story, Gary Shteyngart.  Forget the plot, read it for the social commentary.  The Euro has split into North and South; the only airline left is UnitedContinentalDeltamerican; people think books “smell like wet socks;” teenagers surf each other’s data streams rather than “verballing.”  Welcome to the near distant future.

Be It Resolved: Happy 2013!

Of becoming a writer, Lorrie Morre writes, “First, try to be something, anything, else.”  I get this.  I write because I don’t know how to not write.  Writing helps me understand what I think.  It helps me pay attention.  It helps keep me connected to you.

It’s not always easy to find the time and energy to write.  Despite my intermittent efforts, I haven’t become someone who can spring out of bed at 5am and churn out a blog post every morning.  I get tired.  I get lazy.  I get distracted.

Nor am I someone who finishes every post I start.  Hemingway may be right that one should “write when there is something that you know and not before and not too damned much after.”  But it’s not easy.  I have a file full of half-written posts to prove it.

But I did write in 2012.  37 posts, to be exact.  Less than I’d like but more than nothing.  My most viewed posts in 2012 were:

1. The Impact of Teach For America in Tennessee.  Since writing this post, I’ve gotten another year of data and look forward to updating this post.  Spoiler: Teach For America is still out-performing other teacher prep programs in Tennessee.
2. Gender Matters: Silicon Valley’s Mistreatment of its Daughters.  This piece turned the commendations of a tony Silicon Valley prep school into wordles.  The boys were disproportionately commended for “academic” and “knowledge,” while the girls were commended for “enthusiasm” and “soul.”  A year has passed, and another set of commendations has come out.  If I get my hands on them, I’ll update this post.
3. Minimum Viable Montessori.  This piece compares what Sugata Mitra is building with Hole in the Wall to the principles behind the Montessori method.  Since writing this piece, I’ve gone back and read Montesori’s handbook (written by Maria Montessori herself) and have started tying off the Montessori method to research on how we learn.  I hope to find a reason to write about this in 2013.
4. HBS, HKS, or Both?  It’s not a particularly interesting topic to me, but I understand why it’s of interest to others.  In my file of un-posted blog posts I have two more on this topic.  My guess is that I’ll dust these off and post them the next time I’m scheduled to talk to a prospective student.  Having people read the basics of my viewpoint in advance saves me from feeling like I’m living Groundhog Day.
5. Tune in, Turn on, Don’t Drop Out.  Here I start to make the case for college, which has been getting beat up in the news.  For reasons I’ll share shortly, expect a lot more from me on the value of college in 2013.

Thanks to all who read “A human capitalist” in 2012.  My conversations with you–on and offline–help me see and think more clearly.  I look forward to sharing 2013 with you.

‘Tis Better to Give than To Receive: End of Year Charitable Giving

Happy holidays y’all!

In the spirit that it is better to give than to receive, I made my end of year charitable gifts yesterday.  I thought I’d share a little bit about how I give and to which organizations I give.  I don’t use Charity Navigator or any other charity ratings site.  But I do think about what I value and try to line up my donations with that.

This year, I gave to support:

• The global poor.  More than a billion people live in absolute poverty, none of them in the U.S.  To aid them, I donate to Give Directly, which provides direct financial assistance to those in need.  To put it more plainly, they give money to the poor.  They don’t dictate how it’s spent, but they do research it.  They’ve found that cash transfers benefit children, have a long-term positive impact on household income, and aren’t “misspent” as negative stereotypes of the poor might suggest.  Give Directly currently only operates in Kenya.  But having spent a summer in East Africa, this is just fine with me.

• Services that I value but don’t pay for.  Almost every day, I listen to NPR and look up stuff on Wikipedia.   And neither charges me anything.  To help keep high quality content freely available, I donate to my NPR affiliate, WAMU, and to the Wikimedia Foundation.

• Organizations that have supported me.  I was lucky enough to attend both Duke University and the Harvard Kennedy School on full scholarships.  Out of gratitude, I made small donations to both schools.  Small because these are rich institutions whose students are already afforded so many opportunities.

• Awesome organizations run by friends.  This year, I donated to three organizations in this category: Axium, Gallim, and Turbovote.  Axium Education, founded and run by Craig and Michelle Paxton, provides supplemental math and science programming in rural South Africa.  Their preliminary results are fantastic, and I look forward to writing more about them on my blog next year.  Gallim Dance is a modern dance company based in Brooklyn and run on the business side by Max Hodges.  This year, I saw them perform in New York and Chicago, and I look forward to seeing them perform again in 2013.  Finally, TurboVote, founded and run by Seth Flaxman and Katy Peters, aims to make voting easy.  Thanks to TurboVote, the process of changing my registration and obtaining and submitting my absentee ballot was painless this year.

• Community organizations that drive education reform.  The more work I do in education reform, the more I value the organizations that do the often behind the scenes work of galvanizing interest in and support for improving education.  These are the people who knock on doors, have difficult conversations, attend marathon school board meetings, nudge people to run for school board seats, and track and share performance of school systems on key metrics.  This year, I’m supporting four organizations in this category: Memphis City Schools Foundation, Stand for Children, The Cowen Institute, and the National Math and Science Initiative.  Each of these is also an organization with which I have some connection.

For much of 2012, I worked on the merger of Memphis City Schools and Shelby County Schools.  I believe that their Teacher Effectiveness Initiative is one of the boldest and best led things happening in urban education.  It was in Memphis, for example, that student perceptions—more valid and reliable than student outcomes or a principal’s observation of a teacher—were first used as part of a teacher’s formal evaluation.  Tennessee also has some of the boldest education legislation in the country, for example, tying tenure to teacher performance.  No group worked harder to get this legislation passed than the advocacy group Stand for Children.

The other two organizations go further back.  My first assignment at the U.S. Department of Education in 2005 was to work on the plan for transforming while rebuilding the New Orleans Public School System in the wake of Hurricane Katrina.  Scott Cowen, the president of Tulane, oversaw the planning effort at the same time that he was rebuilding and repositioning Tulane.  I wasn’t the only one taken with his leadership during the aftermath of the storm; a Tulane alum made a significant gift to establish the Cowen Institute.  The Cowen Institute now operates as an action-oriented think tank informing and advancing K-12 education in New Orleans.

I subsequently left the U.S. Department of Education with my boss, Tom Luce, to start the National Math and Science Initiative.  NMSI’s mission is to scale up programs with proven positive impact on math and science education.   There are very few education non-profits that can make clean statements about their impact.  NMSI is one of them.  As the graph below shows, the number of students taking and passing Advanced Placement exams in NMSI supported schools improved from below the national average to almost twice the national average over three years.

How about you?  How do you think about donating to charity?  One change I’ll be making next year is donating earlier in the year.  It’s hard to make an operating budget work when a significant portion of it arrives at the end of December.  But what other changes should I make?  What organizations are on your list?  Or more interestingly, what organizations don’t make your list and why?

Selection Bias Is How Markets Work

Yesterday, I wrote about how a decade of low-road behavior by for-profit universities and supplemental education service providers had left a hangover.  I noted that a major theme of AEI’s #eduprofit event was funding based on outcomes rather than seat time.  Bundled in with that was a plea for greater industry leadership and self-regulation around outcome metrics.

Stacey Childress and Jim Shelton both had ideas for how government and foundations can help identify and scale the most effective products and services—those with the best outcome metrics. Childress, the Deputy Director of Education at the Bill and Melinda Gates Foundation, mentioned that they would be launching a challenge grant for literacy tools.  They hear from educators that they need better literacy tools.  So BMGF wants to create a focal point to attract the market to supply such tools.  As part of their challenge grant, BMGF will pay for the evaluation of effectiveness of such tools so that companies will not have to bear this cost or expend management effort overseeing it.

Shelton, the Director of the Office of Innovation and Improvement at the U.S. Department of Education, suggests that rather than putting in place one time evaluations, we should have a set of lab schools with more advanced instrumentation in place that allows research to be conducted on new products and services without having to build up the evaluation infrastructure each time.  This would be cheaper, faster, and require less overall time on the part of schools.

I love outcomes.  And companies that drive better outcomes.  But I want to raise a flag about these ideas: they both presuppose that evaluators will be better at choosing the most effective tool than customers will.  There are two reasons this is difficult:

1. It is difficult for anyone to know the most important criteria for a product to fulfill upfront.
2. Selection bias is an important part of how most markets work.

It is difficult to know the most important criteria for a product to fulfill upfront.  In the beginning, the e-payments market was crowded.  Some companies set up application processes that effectively screened for fraud but were too onerous to attract many customers.  Other companies made signup and usage easy but got eaten by fraud.  The company that eventually owned this market—PayPal—made signup easy while simultaneously developing advanced algorithms to screen for fraud with minimal information.  It’s easy to say in retrospect that getting both of these criteria right was obviously going to be the winning combination.  But a lot of smart people got it wrong—starting, investing, in, or using the early competitor companies that eventually went bankrupt.  And even PayPal, the company that eventually got it right, didn’t get it right initially.  They started as a platform where it was easy to signup and only when a cofounder started looking into fraud out of personal interest did the company recognize its importance and adapt to it.

The point of this story?  It’s hard to know upfront what the most important criteria are for a product to fulfill.  In the case of a literacy tool, a big part of it is how it improves students’ literacy.  But part of it will also be how the tool respects teachers’ time, how it communicates with parents, how it integrates with other tools, what it costs, and several other things we haven’t thought about yet.  When customers balance all of these criteria and make a purchasing decision, they are sending a signal about what they value.  Companies respond to this, providing more of what customers’ value and less of what they don’t value. Product quality isn’t static.  And it can vary from person to person.

This brings me to the second reason: Selection bias is an important part of how most markets work.  Let’s look at computers as an example.  If we had randomly distributed the first personal computers regardless of people’s willingness to pay for them, we would have quickly concluded that computers were junk.  Chances they would have found a home with someone who could find a use for them would have been low; chances they would have found a home with someone who could find a use worth paying market price for, infinitesimal.  Instead, early entrepreneurs charged market price for computers and those who had a use case bought them.  Companies reinvested this money and made computers that had broader market applicability.  And they did this again and again, until we’ve ended up with netbooks and tablets and smartphones for under $100.

But there is an exception to selection being the predominant way a market should work.  If we believe that engaging in the selection process could significantly harm people, we regulate it.  This is the case for pharmaceuticals, for example.  Companies aren’t allowed to sell prescription drugs that haven’t been through a double-blind randomly controlled trial.  And people aren’t allowed to select into the prescription drugs of their choice.  This is also the case for financial services where companies are required to disclose certain information to the customer before they make their purchase and access to some particularly risky types of products is restricted.

So the question is: are educational products and services more like computers or more like pharmaceuticals and financial services?  And what does this mean for the ideas proposed by Childress and Shelton?  I’ll turn to this question tomorrow.  In the mean time, if you have thoughts, I’d love to hear them.

AEI’s #Eduprofit Event: Old Problems or New Ideas?

Yesterday I attended AEI’s “For-profits and Federal Education Policy” event.  My prior was that the event would be future-oriented. Massive open online courses (MOOCS) from elite universities like Stanford and MIT have attracted hundreds of thousands of students.  Enablers like 2Tor and EmbanetCompass have partnered with universities to attract students and provide student support services such as advising.  And companies like SoFi and CommonBond aspire to upset the higher ed finance market by offering rates below federal lending ones to students at top universities.

Not only is there a lot of innovation taking place but higher education leaders are also under a lot of pressure to wade into the waters of innovation. The University of Virginia’s president was even fired (though subsequently reinstated) for sluggishness in doing so.

Yet, none of these trends were discussed.  Instead, the panel–Stacey Childress,  Raquel Whiting Gilmer, Michael Horn, Jim Shelton, and Eric Westendorf—spent the most time discussing:

1. The roles of the public, non-profit, and for-profit sectors in delivering educational products and services.  The panel was pointed in this direction by moderator Rick Hess, who opened with a question about whether for-profits could put students first given their stakeholders.  This is a fine theoretical question and one that I’ve no doubt thought and care about but feels like old ground at this point.  Also, students are stakeholders…just not the only one, which applies for all of these organization types.
2. The desire of the government to move to outcome-based funding rather than fund based on seat time.  The two examples discussed at length were funding for higher education and funding for supplemental education services (SES), tutoring available under No Child Left Behind to students whose schools don’t make adequate yearly progress.

A decade of low-road behavior by for-profit universities and SES providers has left a hangover that is still top of mind for these panelists, drawing attention away from new ideas and toward old problems.

Comparative higher education finance: the US & Australia

“What can you to say to reassure me…that I will be able to sufficiently support myself after I graduate?”

Jeremy Epstein, 20-year-old exercise science major at Adelphi University asking the first question at the second presidential debate of 2012.

In his response to Jeremy, Governor Romney rightly pointed out that there are two ways to tackle this problem.  One is through job creation.  The other is through the cost of higher education.  Unfortunately, Romney then referenced a Massachusetts program that gives scholarships to students receiving marks in the top 25% on the state’s high school exit exam. While that might relieve some financial burden for top high school students, it is by no means a systemic solution since it offers nothing for the other 75% of students.

Human capital investments are particularly challenging.  For the borrower, it is uncertain whether he’ll be good at the skill he is seeking to acquire and uncertain what skills will be valued in the future.  Unlike in the capital market where an individual can diversify his holdings to mitigate risk, in the labor market an individual has to make employment decisions with long-lasting consequences and limited ability to diversify.  For the lender, it is also uncertain whether the borrower will earn sufficient income to pay back his loan.  But there is the added worry that if he doesn’t, there is no collateral he can seize.  Doing so would at best be indentured servitude; at worst, slavery.

In response to these challenges, the U.S. and Australia have taken different approaches to higher education finance.  In the U.S., colleges and universities set tuition with little price differentiation based on the value of the degree one is seeking.  Jeremy’s exercise science degree would have cost him the same as an economics one, despite the varying value of the degrees.

To assist students in financing their education, the U.S. government offers grants and loans based on financial need.  The loans are mostly non-dischargeable, meaning that they cannot be reduced or eliminated in bankruptcy.  Students can apply for deferrals or grace periods but interest and fees continue to accumulate during this time.

Under the American system, the longer it takes you to pay off the degree, the more the degree costs.  For students from poor families or for those with lower-paying careers this blunts the return they would otherwise get on their degree.  This creates drag on social mobility.

The Australian system looks quite different.  Degrees are priced on a combination of their market value and national priorities.  For example, high-returning medical, law, and economics degrees cost ~1.7x more per term than lower-returning education and nursing degrees.  Australia further sets national priorities—in mathematics, statistics, and science—and prices these national priority degrees even lower than the nursing and education degrees.

Australian students have three options to pay for their degrees.  They can pay upfront tuition at a 10% discount, pay a portion upfront at a 10% discount and the rest in income-contingent loans, or pay the entirety in income-contingent loans.  An income-contingent loan is what it sounds like—a loan whose repayment terms vary with a borrower’s income.

Australian students don’t begin to repay their loans until their income exceeds US$51,135.  If their income dips because they go back to graduate school or can’t find a job or any other reason, they don’t pay anything until their income is back above this threshold.  No fees or interest accumulate during this time.  When income is above the required threshold, repayment ranges from 4% to 8% of income, increasing as income increases.  Because of this, the more money a borrower earns, the faster he pays off his loan.

Australian student loans are indexed to their consumer price index and therefore have a 0% real interest rate.  Yet because of the time value of money, a dollar today is worth more than a dollar tomorrow.  The time value of money means that the longer it takes a borrower to pay off his degree (on a zero real interest loan), the less the degree costs.  Under the Australian system, those taking longer to repay and therefore paying less are those who earn the least.

The outstanding balance of a $25,000 loan if no payments are made. In the US, the loan has a 7% nominal interest rate and a 5% real interest rate. In Australia, the loan has a 0% real interest rate. To simplify, it is assumed that the time value of the money is the same in both countries and equal to 2%. This analysis assumes no fees are incurred.

We are all born with an endowment of human capital that was not of our own making.  We can improve upon it.  But we have to invest to do so, often without knowledge of what the future will look like or of whether we’ll be good at the skill in which we’re investing. A higher education finance system can be agnostic as to career decisions or give clear signals as to the market viability of a degree.  A higher education finance system can make things easier when times are hard or can make things harder when times are hard. Australia’s experience shows that there is more the American higher education finance system can do to answer Jeremy’s question.

–

This essay was first published on NerdWallet. 

Lessons in Stats: Explaining Why Small Schools Are Dumb

I am doing something new in the history of “A Human Capitalist”: shamelessly stealing someone else’s thinking.

Howard Wainer’s piece is just that good.  In “The Most Dangerous Equation,” he gives five examples of how failure to understand that the standard error of the mean is dependent on the size of the sample has led to faulty conclusions with real world consequences.

I want to pull out one of his examples: the movement to break large schools up into small schools.  Those working in ed reform will remember that for years, the Bill and Melinda Gates Foundation (BMGF) championed small schools.  For a time, it was one of their three priorities in U.S. education reform.  And by 2001, Wainer reports that BMGF had made grants of ~$1.7B to push this idea.

Here’s the relevant equation, discovered by Abraham de Moivre: the standard error of the mean = the standard deviation of the sample divided by the square root of the sample size.

The important bit is that the standard error is much larger with a smaller sample size.  With a sample size of 100, we divide by its square root, 10.  With a sample size of 10,000 we divide by its square root, 100.  So there should be more “good” schools and more “bad” schools at a smaller population size.

Wainer takes a look at the data [emphasis mine]:

Shown here at left are math test scores from 1,662 Pennsylvania 5th-grade schools. The 50 highest-performing schools are shown in blue and the 50 lowest in green. Note how the highest- and lowest-performing schools tend to group at low enrollment—just what de Moivre’s equation predicts.The regression line is nearly flat, though, showing that school size makes no overall difference to 5th-grade mean scores. Math scores for 11th-grade schools were also calculated (right). Once again, variation was greater at smaller schools. In this case, however, the regression line has a significant positive slope, indicating that the mean score improved with school size. This stands to reason, since larger schools are able to offer a wider range of classes with teachers who can focus on fewer subjects. 

By failing to understand that small schools with small populations are going to show more variability than large schools, a foundation wasted almost $2 billion of its own money.  Not to mention that it:

1. Wasted the money of school district’s (which is taxpayer money) since small schools are more expensive to operate (think about two small schools with 300 kids that each have their own librarian vs. one with 600 kids that has just one librarian).
2. Put in place policy that was worse for high school students, since math scores rise with school size.
3. Tied up management capacity in converting school sizes when they could have been focused on things that actually matter—like more effective teachers or extended instructional time.

Districts like Boston and Oakland that moved to small schools are now reversing their positions.  Maybe they can use the cost savings to fund the expansion of statistics classes.