Wednesday, January 21, 2015

Nudges and the Parable of Email

“Don’t forget…”  

Susan Dynarski did a brilliant piece earlier this week in the New York Times about “nudging” students towards success.  She draws the term from Cass Sunstein, who wrote a book about behavioral economics and some of its concrete applications.  For example, some schools that wanted students to select fruit more often for lunch discovered that simply placing the fruit in a more conspicuous and convenient place in the lunch line made a difference.  Nudges are consciously-designed interventions intended to take advantage of people’s unconscious habits to push them in a particular direction.  They’re less coercive than directives, in that the nudged are always free to disregard the hint and substitute their own preferences.  But since most people look for the ‘default’ setting on low-stakes decisions, an adjustment to the ‘default’ setting can produce significant results.  Those who care deeply can always override the default, and some will.

Dynarski mentioned several ways that various schools and colleges have used nudges to improve student success.  Her most compelling anecdote, drawn from a randomized study, involves a college using text messages to remind (or nag, depending on taste) students about deadlines to submit financial aid applications.  Apparently, for a cost of about five dollars per student, the college was able to generate a dramatic improvement in rates of students actually completing financial aid applications.  Given the impact of finances on students’ ability to remain focused and succeed academically, that is nothing to sneeze at.  She concluded with a suggestion that nudges don’t catch on because nobody could take internal credit for them.

Concur in part, dissent in part.

I’m a fan of behavioral economics, and I’ve been talking about her article on campus all week.  In the course of some ot those discussions, though, I’ve started to get a sense of some of the friction that new nudges can encounter.

The most basic is a sense of boundaries.  Email provides an example.  Emails are free to send, and every student at the college is issued an email account.  (My college isn’t unique in that; free “.edu” email addresses are pretty standard.)  Students can access email, theoretically, on just about any device.  So over the years, the number of emails that various offices, programs, and departments sent to students mushroomed. When the trickle became a deluge, students stopped trying to sort out the important stuff from the rest, declared it all spam, and ignored it all.  What had started as a potential game-changer quickly made itself irrelevant through overuse.  We had to develop an organization-wide communication protocol with pretty strict rules about mass emails.  Now, when we really need to get responses on a large scale, we use postcards.

I don’t see why texting would be immune to the same issues.  Good results from initial texting would lead to more, and more, until students start to tune it out.  Worse, many of our students are on very cheap prepaid plans that charge by the text, so the barrage of messages actually costs them money.  That may not be true at more affluent campuses, but at most community colleges, we can’t take for granted that students have unlimited plans.

That sort of effect wouldn’t show up in a small, randomized, controlled study.  It’s the kind of thing that shows up with scale, over time.  But it matters.

Identifying a nifty new nudge is great, and I’m all for it.  Figuring how to adapt it to local context in a way that won’t inadvertently kill it is harder.  I’m all for that, too, but it’s a more complicated process.  It’’s not just about claiming internal political credit.  

Having been through that with email, I can understand a certain reticence to adopting wholesale texting as The Next Big Thing.  Any tool is only as good as the strategy behind its use.  It’s all well and good to tell students not to forget about deadlines, but we have to be sure not to forget the mistakes of our past.