Behavioral interventions geared at reducing poverty and inequality are inherently controversial. After all, if the problem is that Group A has a lot more money than Group B, why not give members of Group B more money? Why not attack the problem directly? So it’s no wonder that many of these interventions have been criticized for “blaming the victim,” for trying to “fix” poor people rather than address the system that caused them to be poor in the first place.
On the other hand: People who grow up in impoverished settings may well develop behaviors and habits that make it harder for them to succeed in the real world — not because of any personal failing on their part, but because everyone is deeply influenced by their environment. Poverty and inequality aren’t going away, so don’t they deserve a shot, too? The proponents of anti-poverty behavioral interventions often defend them on harm-reduction grounds: We need to deal with the world we have, not the world we want, and therefore should take a multipronged approach to dealing with these issues.
This debate and its many offshoots are constantly raging in one form or another. One of the most interesting manifestations — interesting in part because it’s so far from being resolved — has to do with the so-called “word gap” between young children who are rich and poor. The basic idea is that, all else being equal, parents from higher socioeconomic classes talk to their kids more than poorer ones, and the more words a kid is exposed to, the better prepared they will be to perform well in school and to operate at a high cognitive level more generally — being exposed to a steady stream of adult speech seems to pave the way for various vital cognitive abilities. Therefore, the thinking goes, if poor families could be taught to talk more often to their kids, and in more sophisticated ways, it might help close the achievement gap.
To some researchers, this is an extremely promising way to close the achievement gap. To others, it’s an overblown, paternalistic idea that has, at its core, a lurking disdain for the way poor and working-class families talk. Either way, the debate itself is a fascinating example of the fireworks that sometimes ensue when behavioral science and social-justice concerns associated with poverty collide.
The best journalistic account of the word-gap theory and its application to real-world settings comes from a 2015 New Yorker story in which Margaret Talbot examines a program called Providence Talks, “the most ingenious of several new programs across the country that encourage low-income parents to talk more frequently with their kids.” Providence Talks, which is ongoing, has parents wear recording devices that allow caseworkers to track how they talk to their kids and give them tips about how to do it in a more enriching manner.
As Talbot notes, the biggest driver of this theory is a landmark 1980s study by the researchers Betty Hart and Todd Risley. “Hart and Risley recruited 42 families: 13 upper, or ‘professional,’ class, 10 middle class, 13 working class, and 6 on welfare. Each family had a baby who was between 7 and 12 months old. During the next two and a half years, observers visited each home for an hour every month, and taped the encounters. They were like dinner guests who never said much but kept coming back.” The researchers eventually analyzed 1,300 hours of recorded conversation, and found that, as they put it, “With few exceptions, the more parents talked to their children, the faster the children’s vocabularies were growing and the higher the children’s IQ test scores at age 3 and later.” This is the origin of the “30 million words” meme and of programs like Providence Talks — by age 4, Hart and Risley calculated, a poor child has heard 30 million fewer words than a rich one. If kids were exposed to more and more varied language, the thinking goes, they’d be better equipped to do well in school when they arrive there.
This has become a very controversial notion. Talbot quotes some researchers arguing that the original finding has been presented in a somewhat overblown and oversimplified way. For one thing, some proponents of the word-gap theory have sometimes placed too much of an emphasis on quantity over quality. And as many researchers have pointed out, a correlational correction between word exposure and intelligence or achievement doesn’t necessarily imply a causal one. Moreover, the Hart and Risley study looked at a grand total of just a small handful of poor, black families — and in one cultural context.
But some of the strongest arguments against the word-gap theory go well beyond methodological concerns. The latest iteration of the anti-word-gap argument comes in OZY, where Leah Durán, an assistant professor at the University of Arizona who studies language, literacy, and young children, both reiterates some of these empirical points, but also sums up a broader argument that pops up frequently: “There is no evidence that if poor parents say more words to their children, those children will do better in school,” she writes. “But anti-poverty measures have demonstrated meaningful and lasting effects on children’s educational outcomes.” Why waste money on a questionable program in which poor parents wear fancy recording gadgets, in other words, instead of just giving them money to help them be less poor? After all, poverty itself can drive the word gap — as Talbot puts it, “When daily life is stressful and uncertain and dispiriting, it can be difficult to summon up the patience and the playfulness for an open-ended conversation with a small, persistent, possibly whiny child.”
These and other arguments have fueled the belief, fairly common in some corners of education and anthropology, that the focus on the word gap is more about politics than about science. As Talbot explains:
Others charged that Providence Talks was imposing middle-class cultural values on poorer parents who had their own valid approaches to raising children, and argued that the program risked faulting parents for their children’s academic shortcomings while letting schools off the hook. Nobody contested the fact that, on average, low-income children entered kindergarten with fewer scholastic skills than kids who were better off, but there were many reasons for the disparity, ranging from poor nutrition to chaotic living conditions to the absence of a preschool education. In a caustic essay titled “Selling the Language Gap,” which was published in Anthropology News, Susan Blum, of Notre Dame, and Kathleen Riley, of Fordham, called Providence Talks an example of “silver-bullet thinking,” the latest in a long history of “blame-the-victim approaches to language and poverty.”
Here are some of the interesting aspects of this debate:
The word-gap debate is a flashpoint for “Why not address poverty more directly?” arguments. Many critics of word-gap interventions argue that part of the reason for the word gap is poverty itself: If parents are endlessly stressed out by the daily grind of trying to make ends meet, then naturally they’re going to have less time for rich interactions with their children. “Between 1994 and 1998, the New Hope Project in Milwaukee, Wisconsin, for example, showed that modestly supplementing poor parents’ incomes led to long-term improvements in their children’s educational outcomes,” writes Durán in her OZY article, for example. “If the real problem is poverty, why not alleviate poverty? Poverty, after all, presents real challenges to children’s learning, such as food insecurity. Parents who need to work two or three jobs to stay afloat have less time to spend with their children, and less money to spend on things like books or tutoring.”
Kathryn Hirsh-Pasek, a developmental psychologist and language-acquisition expert at Temple University, said she strongly disagreed that anti-poverty programs could address outcomes in the same way word-gap programs do. “It’s very simple — just put an exclamation point after the ‘No!’” she said. “Here’s why: In many cases, the parents in these situations don’t know what to do. So you could dump money on it — and we’ve done this, by the way, on school systems, and it doesn’t raise outcome scores. If you don’t know what to do with kids in a way that is consistent with the science, dumping money isn’t the [best way]. And we know that even some middle-class families don’t even appreciate how much you can talk to an 8-week baby. So if you don’t know that, you’re not going to make a choice to do that. This is a story about changing attitudes, knowledge, and behavior.”
In theory it doesn’t have to be either/or, of course, but the problem is that there are finite resources, unfortunately, for programs trying to help low-income people in the U.S. It’s hard to disagree with Durán’s assessment in OZY that “Schools that serve low-income communities are usually under-resourced, and money does matter in efforts to improve educational outcomes in these schools.” James Morgan, a developmental psychologist at Brown University, framed things a little bit differently: “On the one hand, I don’t want to throw cold water on attempts to design interventions,” he said. “On the other hand, I am a little bit concerned — particularly in the case of Providence Talks, because there is a lot of money involved in Providence Talks — that the intervention was designed prematurely.” So it could be the case that some sort of word-gap intervention program could yield long-term benefits, and that anti-poverty programs do too, and that the early iterations of word-gap programs don’t provide much bang for the buck. Again, this is complicated: That’s why the debate isn’t going away.
The word-gap debate shows how deeply we crave simple human-behavior sound bites. Hirsh-Pasek said that while there’s merit to the underlying idea of the word-gap research, it’s been presented in an exceedingly oversimplified way. “I think part of it is that we live in a sound-bite society, and I think nowhere have we ever seen this more than in our political system today,” she said. “And it’s not clear to me as something as big as changing a kid’s trajectory of development for their life is going to be something that’s going to fit into a little tweet.”
“We can show that language is the single best predictor of later outcomes for children all the way through elementary school,” said Hirsh-Pasek. So if there were a way to improve a kid’s language skills, it could make a big impact. But she said the “30 million words” meme has proven so compelling that it has obscured the bigger story, which is that the quality of parent-child interactions matters a great deal: “Everybody ran with the 30 million word gap,” she said. “It was big news. But that piece ignores the quality story, and those of us who have looked at the quality story see quality coming out in many different shades.”
There’s serious debate among researchers over what has been proven at this point about word-gap interventions. Hirsh-Pasek is convinced that researchers have come up with solid interventions to improve what she calls the “duet” between parent and child — the fostering of substantive, complex interactions. Boiling that duet down to the quantity of words spoken misses the point, she said, but the fact is that there is some solid evidence that the interventions themselves do lead to better academic outcomes.
Morgan was less sure. “There are simple ways to measure word count these days, and some interventions that seem to have short-term effects,” he said. “But that doesn’t really mean these are causal in nature, nor that these interventions would have long-term effects — none of these interventions that I’m aware of, and I’ve really been in the field long enough to be able to study long-term effects.”
Sometimes these debates hinge on unanswerable questions about culture and values. Much of the criticism of the word-gap line is hinged on the idea that it “blames” poor children and their parents for their poverty, or that it privileges a certain set of upper-middle-class white parenting and speech mores over others. This argument pops up in many different ways. Some researchers, for example, have cited the fact that in certain Latino cultures, there’s simply less verbose interactions between parents and children in general, and have argued it’s paternalistic to suggest one way of interacting with kids is “better” than another.
When I mentioned this argument to Hirsh-Pasek, she said she disagreed with the idea that it’s inherently offensive to instruct a family on how to interact with their little ones in a way that might help unlock more of their potential. The families themselves, she insisted, are often quite receptive. “I will tell you Latino parents are on the whole so interested in helping their kids make it. It’s just fabulous to watch,” she said. “You have more two-parent families than you have in other communities, you have families who are deeply invested in their kids’ education. And when they learn that this is important, then their kids are going to do better on the other end.”
This raises interesting questions: If an academic says a given intervention is offensive or culturally insensitive, but some members of the targeted group appreciate it and participate in it enthusiastically, whose view should win out? And what if it is a bit culturally insensitive, in the sense of asking people to change deeply ingrained speech norms, but it also has been shown to help? What then? Questions about identity often lack easy answers, and the word-gap debate is no exception.