The impact of a monthly unconditional cash gift on low-income US mothers’ child-directed speech

• In Event: Baby’s First Years: Early Findings from a Poverty Reduction Intervention in Early Childhood

Abstract

The “30 Million Word Gap” first reported by Hart & Risley (1995) sparked a surge of research on language input to children across socioeconomic status groups over the past 25 years. Indeed, studies show that infants from socioeconomically disadvantaged families tend to hear less child-directed speech in their everyday environments than do their more advantaged peers (Dailey & Bergelson, 2022). What drives this difference? Research to date has been unable to specify the causal impact of direct poverty reduction on children’s language input.

We assess the causal impact of monthly, unconditional cash gifts on child-directed speech and child vocalizations among a large, diverse sample of low-income mothers during a 10-minute free play session. We use data from a subsample of the Baby’s First Years study (BFY; Noble, et al., 2021), a randomized controlled trial which recruited 1,000 mothers experiencing poverty in four US cities, shortly after giving birth. Mothers were randomized to receive a monthly, unconditional cash gift of either $333 (“high-cash gift”) or $20 (“low-cash gift”) for the first several years of their children’s lives.

Mothers and their infants (high-cash gift n = 246; low-cash gift n = 325) participated in a 10-minute play session during a home visit at age-1. Using the Language ENvironment Analysis (LENA) system, we extracted estimates of Adult Word Count (AWC), Conversational Turn Count (CTC), and Child Vocalization Counts (CVC) from these 10-minute sessions. These LENA estimates were validated against manual transcriptions of a subset of recordings (¬n=51): LENA counts were significantly correlated with counts from manual transcriptions (Spearman’s r = 0.72 –0.80, ps<.001), similar to previously published rates using longer recordings (Gilkerson, 2017; Cristia, et al., 2020).

To assess the causal impact of the cash gift on child-directed speech and child vocalizations, we conducted intent-to-treat (ITT) analyses on our three LENA estimates: AWC, CTC, and CVC. Our statistical analyses control for child age in months and baseline characteristics (see Table 1 for full list of covariates). We find no statistically detectable impacts of the cash gift on LENA estimates, with or without covariates. See Table 1 and Figure 1.

Follow-up exploratory analyses show statistically significant associations of LENA estimates with maternal education; frequency of parent-child activities (such as book reading); and child language development scores.

In sum, we do not find evidence that the BFY monthly, unconditional cash transfer directly affected low-income mothers’ speech to their one-year-old infants, as measured during a 10-minute free play session. We note that this analysis reports findings after the first year of a pre-registered multi-year intervention, and that even with the high-cash gift, the majority of families remained under 200% of the federal poverty line (Gennetian, et al., under review). Further, we do not have information about the naturalistic language environment over multiple hours, as LENA is typically employed. Ongoing work will assess the impact of the continued cash gift on children’s language input and development over time.

Authors

• Shannon Dailey, Duke Univeristy
  Presenting Author

• Lisa Gennetian, Duke University

• Katherine Magnuson, University of Wisconsin - Madison

• Greg J Duncan, University of California, Irvine

• Sarah Halpern-Meekin, University of Wisconsin-Madison

• Hirokazu Yoshikawa, New York University

• Nathan A Fox, University of Maryland - College Park

• Kimberly G Noble, Teachers College, Columbia University