Category Archives: networks

The how of social capital

Flickr/drjausSocial capital is a powerful resource for individuals and communities.  For individuals embedded in dense social networks, these networks and the attendant norms of trust and reciprocity strongly shape individuals’ ability to land jobs, earn higher salaries, and be happier and healthier.  But, even for those not in the networks, having neighbors who know and trust one another affords benefits in some domains:  better performing local government, safer streets, faster economic growth and better performing schools, among other public goods.

For sure social capital can be used toward negative ends: Al Qaeda, the Crips and the Bloods, the Michigan Militia are all examples where group members can accomplish things that they could not accomplish individually (because of  group social capital).  That said, the literature supports that the vast majority of what social capital is used for is to produce positive ends, not negative ones.

But why?  What makes social capital so powerful?

Robert Putnam and I had always focused on information-flows as the key mechanism.  So these social networks:

  • enable individuals to access valuable information: how to get something done, hear of  job leads, learn how better to promote one’s health, find out what is happening in a community, etc.; or
  • help individuals find partners for joint economic transactions (e.g., to know with whom to partner  in business, to close a sale to a friend or a friend of a friend, to locate a neighbor with whom one can exchange tools or expertise); or
  • spread reputations of members (or neighbors or local merchants) which causes all people in these networks to behave in a more trustworthy manner and facilitates altruism.  There is always a short-term gain to be had from cheating someone, but if the social networks quickly spread the information that one cannot be trusted, this short-term gain is swamped by the lost future opportunity to do business with others; thus it becomes more rational to be honest and trustworthy in communities (physical or otherwise) with strong social networks. Individuals are also likely to be kinder and more altruistic toward others because they know that “what goes around comes around” in densely inter-connected networks and communities; and
  • facilitate collective action: it is easier to mobilize others around some shared goal like politics or zoning or improving trash pick-up if others in the  community already know and  trust you, rather than your having to build those social relationships from scratch.

But Connected (by Nick Christakis and James Fowler) raises a different frame for thinking about this issue: network effects or contagion.  Are there properties of the networks themselves that help spread practices, independent of the flow of information?  This is difficult to answer fully since much of their evidence comes from the Framingham Heart Study where  they know who people’s friends are but not what they are doing with each other or what they are saying to each other.

That said, some of their results can be explained by information flows (e.g., political influence, or getting flu shots), but some seem likely to be working through other channels and not through information-flows (e.g., happiness or loneliness cascades).

In these “network effects” or contagion, Fowler & Christakis typically find that the strongest “network” effects are directly with one’s friends (one degree of separation), but these effects also ripple out two more levels to  friends of one’s friends (two degrees) and friends of the friends of one’s friends (three degrees).  As one would expect, much like a stone dropped in a pond, the ripples get smaller as one moves out.  In fact they refer to the “Three Degrees of Influence” Rule that effects are typically only seen up to three degrees out and not further: in the spread of happiness, political views, weight gain, obesity, and smoking.  For example, in happiness, if one is happy, one degree out (controlling for other factors), one’s friends are 15% happier, at 2 degrees of separation they are 10% happier, and they are 6% happier at 3 degrees of separation.  For obesity, the average obese American is more likely to have obese friends, one, two and three degrees of separation out, but not further.  Quitting smoking has diminishing effects out to three degrees.  For political influence, they note a “get-out-the-vote” experiment that shows that knocking on a stranger’s door and urging the resident to support a recycling initiative had a 10% impact on his/her likelihood to vote for the initiative; what was noteworthy to Christakis and Fowler is that the door-knocking made the spouse (who was not at the door) 6% more likely to support the recycling initiative based on communication with his/her spouse.  They conjecture that if this 60% social pass-through rate of political appeals (6% for spouse vs. 10% for person answering door) applied to one’s friends and if everyone had 2 friends, then one person urging friends to vote a certain way would have a 10% impact on one’s friends, a 6% impact on one’s friends’ friends (2 degrees) and a 3.6% impact 3 degrees out.  Multiplying these political effects all the way through, one vote could create a 30x multiplier. [The example is eye-opening and suggests that voting and political persuasion may be less irrational than thought, but also is based on a huge number of assumptions and assumes no cross-competing messages from friends.]

In an experiment on altrusim (explained in this post) Christakis & Fowler found that $1.00 of altruism, ultimately produced $1.05 of multiplier effect ($.20 one ripple out with 3 others and $.05 of altruism two ripples out with 9 others).

Christakis and Fowler, in their book, talk about contagion effects in voting, suicide, loneliness, depression, happiness, violence, STDs, number of sexual partners, binge drinking, back pain, and getting flu shots, among others.  [One summary of many of their findings, which they note, is "You make me sick!"]

Why do these effects only reach out 3 degrees of influence?  Christakis & Fowler suggest 4 potential explanations.

1) intrinsic decay: C&F liken this to a game of telephone where as the information gets repeated, the content gets lost, or the passion and knowledge of the initiator gets dissipated.

2) Instability of ties: because of what is known as “triadic closure“, if A is friends with B and B is friends with C, it is likely that A will become friends with C.  Because of this, closer-in ties between people have more routes connecting them, and further out ties are more dependent on only one pathway connecting them.  For example, assume Abby and Fran were friends 3 degrees removed via Bert and via Charles. If any of these intervening friendships end (say Bert is no longer friends with Charles), Abby loses her tie to Fran.  Thus, these outer ties are much less stable and averaged across all the “3 degrees of influence” friendships, many more may have zero effect because the path of influence dies out as friends change.

3) cross-information:  as one gets further out away from you, say the friends of the friends of your friends, all of these folks are getting lots of cross-stimuli from lots of other sources (many of which may come from different clusters with different habits or values) and these cross-stimuli start to cancel each other out.

4) evolutionary biology: C&F note that humans evolved in small groups that had a maximum of three degrees of separation so it may be that we became more attuned to being influenced by folks who were in a position to alter our gene pool.

So what are the network influences independent of communication.  There seem like 6 possible channels, and often it is hard to separate one from the other, although some may make more sense for the spread of behaviors and others may make more sense for spread of attitudes or emotions:

1) homophily: “Homophily” is the practice of befriending others like you — “birds of a feather flock together.” Being friends with people who are different than you can be stressful.  This is why in mates and in friends we are likely to choose others with whom we have a lot in common — think of arguments you’ve had with friends about where to go for dinner or what is right or wrong with the world when those friends have very different tastes or politics.  For this reason, one reason for increased clustering over time of obese people or smokers or binge drinkers is that it is stressful to be in groups where one is the minority and either constantly noodging others to change their behavior or else your finding yourself frequently doing what your friends want to and what you do not (e.g., eat fast food, smoke, or listen to heavy metal rock music).  As a consequence, people may vote with their feet and form new ties or strengthen ties with others with whom they have more in common.

2) norms/reference groups/culture/peer pressure:   we often measure the reasonableness of our behavior against our friends.  For example, if our teen friends have all had 6 sexual partners in the last year, then repartnering seems far more normal than if one is friends with a group that is heavily monogamous.  Ditto with obesity or smoking or other possible traits or behaviors.

3) subconscious/imitation:  as suggested with “emotion” below, sometimes we mirror others’ behavior or emotions without even thinking about it.  C&F say it makes sense to think of people as subsconsciously reacting to those around them without being aware of any larger pattern.  They talk about processes by which a “wave” at a sporting event takes place, or fish swim in unison, or geese fly in a V-formation, or crickets become synchronized — all of these happen by individuals mirroring those around them.  And in the process, emergent properties of the group arise (much like a cake takes on the taste unlike any of its individual ingredients).

4) emotions: C&F note that emotions actually affect our physical being — our voices, our faces, our posture.  In experiments, people actually “catch emotions”: others become happier by spending time around happy people or sadder by hanging out with depressed individuals.  In experiments, smiling waiters get bigger tips.  It seems quite plausible that cascades like loneliness, happiness, depression, etc. could spread simply from emotional states, independent of any information flowing through these friendships.

5) social invitations for shared action: friends often invite friends to do things — that’s part of friendship. For behaviors, one of the ways they can spread through networks is that, for example, thin friends could invite friends to exercise more, or obese friends could encourage friends to get ice cream together, or smokers could encourage others to leave the dance for a cig.

Connected notes that it is often hard, for example, to tell imitation and norms apart, “When a man gives up his motorcycle after getting hitched, is he copying his wife’s behavior (she doesn’t have a motorcycle) or adopting a new norm (the infernal things are unsafe?)”

Connected also notes how behaviors or attitudes can spread several social links out, even without the intervening link changing.  They suggest that Amy could have a friend Maria who has a friend Heather.  (Amy and Heather don’t know one another.)  Heather gains weight.  Maria, who really likes Heather, becomes less judgmental of her weight and gradually less judgmental of  obesity in general.  Maria doesn’t change her behavior but when Amy stops exercising with Maria, Maria is less likely to pressure her to resume.  Thus Heather’s obesity changes Amy via Maria (by Maria no longer urging her to keep exercising), but Maria doesn’t change her behavior and Amy and Heather don’t know one another.

It’s interesting stuff to ponder and makes one think more expansively about the role and mechanisms of social capital.  It also evokes a conversation with a Saguaro Seminar participant back in 1998 concerning whether black kids and white kids doing sidewalk painting together on the steps of an art museum could promote inter-racial trust, even if the black kids and white kids didn’t know each other, didn’t talk to one another and never met again.  [My hunch is yes, depending on the strength of their pre-existing beliefs about inter-racial trust, but that talking could make the exchange far more powerful.] Another Saguaro participant wondered whether singing together in a chorus helps build social capital, even if one never has a conversation directly with another member of the chorus.  (In the latter example, in addition to being highly unlikely, you are at least getting some non-verbal information over time from the other choral members about their trustworthiness: do they come regularly and on time, do they respectfully listen to and follow the choralmeister?)

I welcome your thoughts.

For more on the network effects, read pp. 24-30, 25-43 and 112-115 in Connected.

Best friends may provide the most new and valuable info

Flickr photo of BFF from nokapixel

Mark Granovetter in a famous 1973 article “The Strength of Weak Ties” observed that it is our weaker social ties that are most likely to provide us access to information we don’t already know about: job leads, cross-fertilizing information that we can use to great advantage in our jobs, new opportunities, etc.

A recent paper by Sinan Aral and Marshall Van Alstyne says that Granovetter neglected to include frequency of contact.  Yes, our weak ties are more likely per contact to provide us new information, but we contact our strong ties so much more often that a majority of novel information actually comes through those strong and demographically similar friends.

Aral and Van Alstyne analyzed nearly a year of e-mail from an executive recruiting firm (heavily dependent on e-mail for communication and where novel information was critical in finding the right candidates) and found that those with a tighter group of friends (which they define via less network diversity) actually got a higher ratio of new information per unit of time and produced higher revenue for the firm.  As the authors hypothesized, recruiters with more diverse networks suffered a big drop in the volume of communication (what they call “channel bandwidth”).  “Interestingly however, reductions in channel bandwidth associated with greater network diversity do not seem to be driven solely by time and effort costs of network maintenance, but also by the nature of the relationships in sparse networks.” Van Alstyne concludes:  “a smaller number of high-bandwidth relationships can be good for you.”

So no need to jettison your best friends for now…

Read “Buddy System” (WIRED May 2011, by Clive Thompson)

See Sinan Aral and Marshall Van Alstyne, “Networks, Information & Brokerage: The Diversity-Bandwidth Tradeoff” (2010)

The Friendship Paradox: using social networks to predict spread of epidemics

Nick Christakis and James Fowler (whose research we’ve previously highlighted) is back with research that shows how one can easily use “sensors” in a network to track and get early warning regarding the spread of epidemics.

They took advantage of the “friendship paradox” to do so.  In any real-life network, our friends are more popular than we are.  [This is true mathematically in any group with some loners and some social butterflies.  If you poll members in the group about their friendships, far more of those friends who are reported are going to be the social butterflies.  If far more people reported friendships with the loners, they wouldn't be loners.  See discussion here.]

Thus by asking random people in a network, in this case Harvard students, about their friends, researchers know that their friends are more centrally located in these networks.    Then one can track behavior among the random group and their friends, in this case the spread of H1N1 flu (swine flu) among 744 Harvard students in 2009.

Those more central in these networks (the “friend” group) got the flu a full 16-47 days earlier than the random group.  Thus, for public authorities, monitoring such a “friend” group could give one early indication of a spreading epidemic; they could serve as “canaries in the coal mine”.  If the process of spreading was person-to-person rather than being exposed to some impersonal information (via a website or a broadcast), one could also track the difference between a random group and a friend group to predict other more positive epidemics, like the spread of information, or the diffusion of a product, or a social norm.

We write in general on this blog about the positive benefits of social ties (social capital), but Fowler and Christakis’ study also shows you that having friends and being centrally located has its costs: in this case getting the flu faster.  [In some ways, this is analogous to Gladwell's discussion in the Tipping Point of how Mavens, Connectors and Salesmen may be disproportionately influential in the spread of ideas through networks, although Fowler and Christakis are far more mathematical in identifying who these central folks are.]

The “friends group manifested the flu roughly two weeks prior to the random group using one method of detection, and a full 46 days prior to the epidemic peak using another method.

‘We think this may have significant implications for public health,’ said Christakis. ‘Public health officials often track epidemics by following random samples of people or monitoring people after they get sick. But that approach only provides a snapshot of what’s currently happening. By simply asking members of the random group to name friends, and then tracking and comparing both groups, we can predict epidemics before they strike the population at large. This would allow an earlier, more vigorous, and more effective response.’

‘If you want a crystal ball for finding out which parts of the country are going to get the flu first, then this may be the most effective method we have now,’ said Fowler. ‘Currently used methods are based on statistics that lag the real world – or, at best, are contemporaneous with it. We show a way you can get ahead of an epidemic of flu, or potentially anything else that spreads in networks.’

Christakis also notes that if you provided a random 30% in a population with immunity to a flu, you don’t protect the greater public, but if you took a random 30% of the population, asked them to name their friends, and then provided immunization to their friends, in a typical network the “friend” immunization strategy would achieve as high immunity protection for the entire network as giving 96% of the population immunity shots, but at less than 1/3 the cost.

The following video shows how the nodes that light up first (markers for getting the flu) are more central and far less likely to be at the periphery of the social network.  The red dots are people getting the flu; the yellow dots are friends of people with the flu and the size of the dot is proportional to how many of their friends have the flu.

Good summary of this research and its implications here: Nick Christakis TED talk (June 2010) – How social networks predict spread of flu.  Nick also discusses some of the implications of computational social science, which we’ve previously discussed here under the heading of digital traces.  Nick discusses how one could use data gathered from these networks (either passively or actively) to do things like predict recessions from patterns of fuel consumption by truckers, to communicate with drivers of a road of impending traffic jams ahead of them (by monitoring from cell phone users on the road ahead of them how rapidly they are changing cell phone towers) to asking those central in a mobile cellphone network (easily mapable today) to text their daily temperature (to monitor for impending flu epidemics).  Obviously these raise issues of privacy, which Nick does not discuss.

News release of study

Academic article in PLoS ONE

James Fowler on The Colbert Report discussing the book by Fowler and Christakis called Connected.


Nick Christakis presenting a talk at TED — The Hidden Influence of Social Networks. (February 2010).  In the talk he notes that while almost half of the variation in our number of friends is genetically-based (46%), that another equally large portion (47%) of whether your friends know each other is a function of whether your friends are the type that introduce (“knit”) their friends together or keep them apart (what they call “transitivity”).  About a third of whether you are in the center of social networks or not is genetically inherited.  Christakis believes that these social networks are critically important to transmitting ideas, and kindness, and information and goodness; and if society realized how valuable these networks were, we’d focus far more of our time, energy and resources into helping these networks to flourish.

Happiness is contagious

dancingfriendsNick Christakis (Harvard School of Public Health) and James Fowler (Univ. of Calif., San Diego), who previously used the Framingham Heart Study to show that having fat friends increasingly makes people obese, are back with a very interesting paper showing that happy friends make you happy — what the co-authors called ‘an emotional quiet riot’.

It is already established that happiness and having social capital (friendships) are linked, but this research demonstrates that it matters how happy your friends are and that it is the happy friends that are causing your happiness rather than vice versa. Conversely, having unhappy friends over time makes you less happy.

The research shows up in the latest issue of BMJ. “Dynamic spread of happiness in a large social network: longitudinal analysis over 20 years in the Framingham Heart Study.”  [The study involved 5,124 adults aged 21 to 70 who were followed between 1971 and 2003.]

They measured happiness with a 4-item construct:  “I felt hopeful about the future”; “I was happy”; “I enjoyed life” and “I felt that I was just as good as other people.”

They found that happiness is a network phenomenon, clustering in groups of people that extend out to 3 degrees of separation (the friends’ friends of one’s friends), but with greater impact on friendships that are 2 or 1 degree of separation from you.  Demonstrating the magnitude of this effect, co-author James Fowler noted, “if your friend’s friend’s friend becomes happy, that has a bigger impact on you being happy than putting an extra $5,000 in your pocket.”

They found that happiness spreads across a diverse array of social ties, from spouses to siblings to neighbors. They found no happiness effect of co-workers and found that nearby ties had a far greater influence than distant ties: for example, knowing someone who is happy, makes you 15.3% more likely to be happy, but having happy next-door neighbors makes you a full 34% more likely to be happy (much higher than having happy neighbors merely on your block). The optimal effect was for a happy friend living less than half a mile away, which boosts your chance of happiness by42%. In one of the study’s surprises, happy spouses (which one assumes live less than a half mile away!) only increased one’s chance of happiness by 8%. Part of the lower spouse effect is that happiness spreads more effectively through same sex relationships than relationships (romantic or not) between a man and a woman.  (Gays take note!)  Christakis and Fowler believe we may take emotional cues from people of our gender.

They observed that network characteristics (where you were in the network and how happy the people were around you) could independently predict which individuals would be happy years into the future.

They suggest that there may be an evolutionary basis for human emotions.  Previous work noted that emotions like laughter or smiling seemed evolutionary adapted to helping people form social bonds.  [They note: "Human laughter, for example, is believed to have evolved from the 'play face' expression seen in other primates in relaxed social situations. Such facial expressions and positive emotions enhance social relations by producing analogous pleasurable feelings in others."]

While they couldn’t prove it, they suggested 3 possible causal mechanisms:

  1. happy people might share their good fortune
  2. happy people might change their behavior toward others (by being nicer or less hostile)
  3. happy people might exude a contagious emotion (although this would have to be over a sustained time period)

Christakis and Fowler noted that the 3-degrees of separation impact observed in happiness was the same as for smoking and obesity (which also reached out 3 degrees). They wonder whether a “3-degrees of influence” extends to behaviors like depression, anxiety, loneliness, drinking, eating, exercise and other health-related activities.

So the next time you’re unhappy realize that you may be “infecting” your friends with unhappiness as well.  Christakis’ work is suggesting that we need friends, but we also need to carefully pick friends that are happy and have healthy behaviors or we risk that their unhappiness and unhealthy behaviors will spread to us.  The New York Times notes that one of the co-authors indicated that he now thinks twice about his mood knowing that it affects others. That said, he noted: “We are not giving you the advice to start smiling at everyone you meet in New York….That would be dangerous.”

While they think that face-to-face connection is important in spreading happiness (hence the decline of these effects with distance), they did a separate study of 1,700 Facebook profiles, where they found that people smiling in their photographs had more Facebook friends and that more of those friends were smiling. While the Facebook study is just an initial foray into the online word, Christakis thinks that it shows that some of these happiness findings might extend on social networking as well.  And it would take longitudinal studies to determine whether our online activities are gradually eroding our need for face-to-face communication to spread happiness.

Note: Justin Wolfers (on the Freakonomics blog) is skeptical of this research.  As he notes:

[It's possible that it is not your friends' happiness that is causing yours, but that "if you and I are friends, we are often subject to similar influences. If a buddy of ours dies, we’ll both be less happy. Or, less dramatically, if our favorite football team wins, we’ll both be happier. But this isn’t contagious happiness — it is simply a natural outcome of the shared experiences of people in the same social circles. Unfortunately, observational data cannot distinguish the headline-grabbing conclusion — that happiness is contagious — from my more mundane alternative: friends have shared emotional influences."

Wolfers notes that a very careful article by Ethan Cohen-Cole and Jason Fletcher uses the same research design to show how it can lead to silly conclusions.  Cohen-Cole and Fletcher find in another dataset that this approach shows “height, headaches, and acne are also contagious.” As Wolfers notes, it’s more likely that “the same jackhammer causing your headache is likely causing mine.” And the height finding is obviously not causal but more likely a function of homophily (people choosing similar friends).

See Clive Thompson, “Is Happiness Catching?” (NYT Sunday Magazine, 9/13/09)

Boston Globe story available here.

New York Times story available here (which also has a nice graphic showing the clustering in this network of happy and unhappy people).

L.A. Times story available here.

See Wolfers’ Freakonomics blog post here.

The Pay-for-Performance NYC schools experiment: the Social Capital Story

Roland Fryer, the up-and-coming Harvard Economics Professor, who at age 30 is on leave as the Chief Equality Officer for the New York City Public Schools, talked yesterday at the Taubman Center on State and Local Government’s Annual Meeting. [Taubman director Ed Glaeser commented on the apparent tension between the words *Chief* and *Equality* in Roland's title, to which Roland noted that he's all in favor of equality as long as he can be the leader!]

Roland has been on a quest to marry economics and social science with how to make a difference in the lives of the very poor. Roland himself comes from a poor family growing up in the South (in Daytona Beach, FL), and to some extent wants to enable more poor kids to achieve what he has. [Fryer noted that his grandmother's sister and her husband, with whom he spent a lot of time. ran a crack ring, and he happened to be at the dog track watching greyhounds when federal agents arrested the two of them. And 8 out of 10 of his closest friends growing up had either died or served time in prison. And he would have gone to jail as he was planning at age 15 to participate with friends in a burglary, but got cold feet and chickened out at the last minute and his friends were all arrested and went to jail.]

His NYC schools experiment aims to marry cutting edge social science (and randomized intervention) with making a difference in these kids’ lives.

The program has a handful of different dimensions including incentive pay for principals, offering cellphones to high school youth who are doing well economically (called *The Million*), paying 4th and 7th graders to take math and reading tests that they were already taking several times a year (with higher pay if they get more right answers on them).

There has always been the most controversy around the *pay for performance* part of the plan, with some principals saying kids should be studying for the love of learning or claiming that when the financial incentive is removed in the future, any patterns of success will disappear. Roland noted that the media (originally negative) has turned positive with support for the plan in the Washington Post, USA Today and a handful of other papers. Interestingly, while white parents on Upper East Side or in Staten Island complaining that the pay-for-performance didn’t respect African Americans and played into stereotypes, there has been no such complaints among African-American parents of students in the pay-for-performance experiment.

Part of the way through the first year, the plan appears to be working for 7th graders, but less so for 4th graders. Roland notes that it is possible that the tests (every 5 weeks) are too far away for the financial incentive to mean as much, or that 4th grade is too young a grade to incent learning, although Roland is likely to try some tweaks to the program in the second year. Moreover, Roland is going to expand the 7th grade program to 8th grade randomizing who continues in the program and who does not (so they can see how the performance of those who do not get the program in the second year compare with those who never got the treatment or those who got it for two years). Among the 7th graders, those who got offered the treatment (pay for performance) were already 1-2 months ahead of their comparable peers after only 1/3 of a year in the program, although gains were higher in math than in verbal skills.

Hearing Roland talk, it was clear (even though he didn’t directly discuss this), the importance of social networks, trust and social capital to his success. He noted for example, that although principals think that they are in charge of these underperforming schools, that it is really probably 20 students who are running the place and his effort aims to change the culture of schools and create a demand-side for learning that can be spread through social networks. It is no accident that as part of *The Million* (his effort to brand student achievement and success), that he gives students state of the art cellphones, since the cellphones promote social networks where Fryer can communicate with the students or send a video (since he has their phone numbers), where teachers can text students about upcoming tests, and where students can spread the excitement of achievement through this program. And he notes that the payment can serve as a valuable foil for legitimzing studying hard; he hears about students claiming that ‘they don’t care about studying, that they are just in it for the money.’ Roland doubts this is true, but it protects them against the charge that African-American kids are “acting white.”

Roland also notes that social networks will play a role in the anticipated higher take-up rate in second year of program. In first year, many students failed to participate. Growing up in a culture and environment where many people couldn’t be trusted, many students thought the program was a scam. But now that students have seen payments deposited in the bank accounts that were automatically set up for them, word-of-mouth is spreading the messages that the program is legitimate.

Roland told a funny story –he has the look of a nerdy African-American with wire rim glasses and a three-piece suit, with a build of someone who might have played football at an earlier age). Yesterday, a group of wide-eyed NYC students had earned a trip to Harvard because of their performance. He asked the group of 7th graders how many of them expected to be professional athletes and 70% of them raised their hands. He then asked them which of them was the best athlete. They pointed to a 6 foot high kid. Roland made the kid a bet; he’d challenge the kid to a 50 yard dash in Harvard Yard. If Roland won, the kids would abandon their goal of professional athletes and focus on making it in academics since they were never going to be professional athletes if the best of them could be beat by a Harvard economics professor. If Roland lost, he’d wear whatever clothes they wanted him to wear for a day (they had decided it would be a frilly pink dress). Luckily for Roland, he won, although he said it was not easy.

Note: all of funding paid for in first year by foundations, with commitment that if the program shows results, NYC will pick up the tab of continuing this going forward.

For a description of Roland see this account by Freakonomics author Steven Dubner in NYT; for some accounts of this program see this NYT article, or “Cellphones as Incentives,” and this entry on “The Million