On the question of targeting your message: Slow jamming the news

I imagine many of you all have seen the clip of our President slow jamming the news on Jimmy Fallon’s Late Night. Specifically, the issue of the interest rate hike about to go into effect this summer on Stafford student loans.

Of course I can’t help but look at it with a framing eye. It answers a great question I often get asked about whether different messages are needed for different audiences.

The content of the President’s message isn’t different from what he’s been saying all along – that we should care about this because keeping college affordable is important for our future, and that it’s in our power to stop the rate hike.

What’s different is how he’s saying it – directly to students on a college campus, and with language, music and humorous analogies that have meaning to them.  He tailored his message, he didn’t change it completely.

And so, a lesson from our President (and, of course, the Late Night writers), on how to think outside the box on message delivery. Check it out: The Stafford loan slow jam

-lynn

What arguing about baseball with my mom taught me about statistics and storytelling

“We don’t base what we believe on the real world, we base what we believe on the image of the world that’s in our minds. When you change the way that the image enters the mind, you change what you believe to be true.”  Bill James, baseball statistician

I always knew that baseball would occupy a special place in my life. I grew up outside of DC with parents who were displaced New Englanders and die-hard Sox fans. When my dad turned 60, my mother gifted him with a trip to Red Sox Fantasy Camp – he still has the uniform, and puts on at least the jersey, number 9, for important Sox games.  While I had visited Fenway as a kid during summer vacations – the years of Yaz, Fred Lynn, Jim Rice – perhaps my fondest baseball memory is when my mom and I went to see the Sox play the Orioles the year the beautiful Camden Yards opened in Baltimore.  We bought a program, kept score, and I listened as she analyzed the line-up, anticipated who’d have trouble against the pitcher and why. Aside from the striking beauty that is Camden Yards, I think I remember this so vividly because my mother, a woman who was almost entirely directed by her intuition, seemed a little data-driven at the ballpark. And I was a little data-driven everywhere.

I was in grad school at the time, had more than my share of advanced statistics courses, and, somewhere along the way, had happened on Bill James’ Baseball Abstract. It fascinated me (if you’re not familiar, but have seen the film Moneyball, Billy Beane’s radical rethinking of the Oakland A’s roster was based on James’ theories and data analysis). And it was there at Camden Yards that I shared with my mom some of the Jamesean analysis that debunked traditional strategies, including the proposition that stealing bases isn’t worth it in the long run.

“You and your data,” she said. “That’s ridiculous. Stealing third in particular. Runners are more likely to score from third than second, so why not steal if the opportunity is there?”  I explained that a player would have to have a crazy high success rate of stealing bases in order for the gain to offset the potential of losing that runner to an out.  She scoffed, said something about players trusting their instincts.  I said, “I’m just saying…the data indicate that on average, over the long run, there’s no relationship between base stealing and wins.” She responded that it had to be worth it, else they wouldn’t do it.

“There’s a great deal of received wisdom, a great deal of things that people have known for generations….and most of the time…the things that people know are the enemies of what they learn.” This I did not say to my mother. It’s just another classic Bill James quote.  And it aligned with everything I was learning in grad school about social cognition. My dear mother was resisting facts that didn’t align with her view of how the game should be played! I pointed that out to her – citing the science to back it up, of course. Appropriately, she scoffed.

And therein lies the communications lesson (you were probably wondering if all this reminiscing was going to lead somewhere productive, eh?) For most people, data do not “change the way the image enters the mind.” We need to start with the meaning of the data, and not assume the data will create the meaning on its own.  We need a data-driven story that doesn’t lead with the data, but leads instead with a hook that taps into those long-held, based-on-decades-of-fandom beliefs that people have.

Knowing this, and wanting to share the wealth of information I’d gleaned from the Bill James statistical pile up in my head, I decided to tell my mom a story about Babe Ruth. Or, more accurately, a story that aligned with her seemingly constitutional desire to deride the Yankees.  See, in 1919, his last season with the Red Sox, Ruth hit what was seen as an unreachable 29 home runs.  But then came 1920, the year Babe Ruth was traded to the Yankees (for cash!), the year he hit 54 home runs (no other player had more than 19). To all Red Sox fans, this was a double-whammy – going to the Yankees and smashing his own Boston-based record  (which, of course, he’d do again and again until reaching 60 in 1927, a record which held for 34 years, give or take an asterisk, when Roger Maris hit 61).  Okay. There I go again with all the data. And I know better.

Anyway, I said to my mom, “You know all that hoopla around Babe Ruth going bananas with homers when he moved to the Yankees? Did you know that he hit 29 homers with the Sox in 1919, and that was considered “unreachable”? Ever thought about why he was able to hit so many more, in a single year?”  She looked at me, interested.  “Did you know that Fenway was a really tough home run park, and only 9 of his 29 homers were at Fenway?” I could see the wheels turning in her head: Fenway, way tougher home run park than the Polo Grounds, of course! “And, did you know that in 1919 the American League played a shortened schedule, but he played a full schedule with the Yankees in 1920?”  “AND…” I continued,  “he was still pitching in 1919; in 1920, when he went to the Yankees, he played right field full-time!” “Well, that explains it!” And her eyes lit up in a way that Red Sox fans’ eyes light up when an answer to a question in baseball isn’t “pinstripes.”

Find the meaning in the data. Relay the meaning, and let the data follow.

A final note: My father and I went to a Red Sox-Yankees game at Fenway last summer and he told me, as I ate my Fenway franks and he, clad in number 9, ate his Italian sausage, that he’d first hitchhiked to a game at the tender age of 11, making this the 8th decade in a row that he’d watched the Red Sox at Fenway (so if you’re wondering where my preoccupation with numbers comes from, well there ya’ go. DNA). During the game, when David Ortiz approached the plate, the Yankees used an infield shift and my dad pointed it out, explaining this modified “Ted Williams shift” to me.  I knew what the Ted Williams shift was – it’s where the defense, when facing a typically left-handed hitter who pulls disproportionately to right-field, e.g., Ted Williams, shifts three infield players to the right side of second base –  but I didn’t have the heart to tell him.  I also knew in some detail the argument that Bill James, now employed by the Red Sox, makes against using the shift. But I didn’t have the heart to tell him that, either. It was something to think he’d seen the actual Ted Williams Shift first deployed by Sox opponents in the late 40s, in decade one of his visits to Fenway. That was enough. THAT was a story.

 

Science News that Advocates Can Use: Known features of welfare programs that improve outcomes for parents and children

When I read a fascinating article in the journal Child Development about welfare program characteristics and both parent and child outcomes, I immediately searched Google news for a news story or editorial that may have attempted to relay these findings to the public. I found none. Such is often the case with scholarship – it is read and discussed by scientists, but often fails to get translated for public consumption.  This gave me the idea to occasionally blog about important science news that hasn’t yet become a public conversation, but which would be of great interest to advocates and policy makers. Of course, one misses a lot of interesting and rich detail in summarizing for a blog space, so please see the original research for the full account. That said, here is the inaugural post of Science News that Advocates Can Use.

 —

Erin B. Godfrey of New York University and Hirokazu Yoshikawa of Harvard University conducted research that explored whether differences in how welfare offices implement welfare-to-work might be linked to parent and child outcomes.  The outcomes they studied were parents’ income and welfare receipt over time, and several important measures of children’s development.

Godfrey and Yoshikawa’s study draws on more than a decade of prior research on welfare policy, and decades of theory and research on both social support and family well-being. Considering parent outcomes, they hypothesized that:

1)   The amount of support provided to parents by caseworkers would improve parents’ earnings and income over time (measures of support included fostering trust, mutual respect, listening and openness)

2)   An office’s caseload size would have an inverse relationship to parents’ earnings and income over time (i.e. larger office caseloads would be associated with lower incomes), and

3)   The degree of emphasis placed by the welfare office on parents’ securing employment quickly would lead to larger increases in parents’ earnings and income over time, and larger decreases in their welfare receipt.

What did they find? Following the above hypotheses:

1)   Recipients with “high-support “ welfare offices had steeper increases in earnings and income over time (they measured over 5 years). The authors also found that caseworker support varied more across welfare offices than across caseworkers, and is thus more characteristic of the policy implementation setting than it is characteristic of the individual caseworker.

2)   Recipients in offices with high caseload sizes had greater decreases in both their earnings and their welfare benefit over time.  The authors note that prior research has shown that caseload size is an important determinant of the quality of relationship between caseworker and recipient – large caseloads may reduce caseworkers’ time with recipients, increase the chance of caseworker errors, and increase the chances that recipients are mistakenly sanctioned for program violations.

3)   Finally, recipients in offices with a strong emphasis on employment had large decreases in welfare receipt, but there was no detectable influence on income or earnings over time. The authors interpreted this to mean that these offices were successful at quickly moving recipients into employment and off welfare, but that the quality of jobs taken by recipients may not have been high.

The authors next set out to determine whether parents’ income over time had any impact on cognitive and behavioral outcomes in their children.

1)   They found that children whose parents experienced steeper increases in overall income over time had significantly higher scores on standardized tests of reading and math.

2)   They also found that children’s withdrawn and depressive behaviors decreased when either parents’ income from work or welfare receipt grew at faster rates.

The authors rightly note that their findings should be of interest to policy makers and service providers who are interested in improving economic circumstances for families and developmental outcomes for children.  They suggest, “Given that states have considerable flexibility in designing and implementing welfare programs (US. Government Accounting Office, 2002), they could design programs that reduce the negative influence of caseload size and foster caseworker support.” (p. 396)

It is always jarring to me to read scholarly research that shows what effective government programs look like (and with very real positive effects on families and children) at the same time that the widespread public story about government is that public programs need to be cut to “battle inefficiencies.”  Godfrey and Yoshikawa’s study adds to our knowledge of what efficiency and effectiveness look like in welfare-to-work programs:  reducing caseload size, ensuring offices foster a culture of supportive interactions between caseworkers and recipients, and being wary of strategies to secure quick employment regardless of the wages those jobs supply.  These features of how welfare-to-work programs are implemented improve the very outcomes we charge the programs with accomplishing, and more: improving families’ economic circumstances, but also the academic progress and mental health of their children.

Godfrey, E.B. and Yoshikawa, H. (2011). Caseworker-Recipient Interaction: Welfare Office Differences, Economic Trajectories, and Child Outcomes. Child Development, 83 (1), 382-398.