Baby's First Years

Recent advances in developmental neuroscience suggest that experiences early in life have profound and enduring influences on the developing brain. Family economic resources shape the nature of many of these experiences, yet the extent to which they affect children’s development is unknown. Our team of neuroscientists, economists and developmental psychologists proposes to fill important gaps in scientific knowledge about the role of economic resources in early development by evaluating the first randomized controlled trial to determine whether unconditional cash gift payments have a causal effect on the cognitive, socioemotional and brain development of infants and toddlers in low-income U.S. families. Specifically, 1,000 mothers of infants with incomes below the federal poverty line from four diverse U.S. communities will receive monthly cash gift payments by debit card for the first 40 months of the child’s life. The high cash gift treatment group mothers (40% of all mothers) will receive unconditioned cash payments of $333 per month ($4,000 per year) via debit care for 40 months. Mothers in the low cash gift comparator (control) group (60% of all mothers) receive a nominal payment – $20 per month, delivered in the same way and also for 40 months. The $333 per month is an amount in the range of a variety of income assistance policies in the U.S. and has been shown to be associated with meaningful improvements for poor children in prior studies. In order to understand the impacts of the added income on children’s cognitive and behavioral development, we will assess high vs. low cash gift group differences at age 3 (and, for a subset of measures, at ages 1 and 2) on measures of cognitive, language, memory, self-regulation and socioemotional development. Brain circuitry may be sensitive to the effects of early experience even before early behavioral differences can be detected. In order to understand the impacts of added income on children’s brain functioning at age 3, we will assess, during a lab visit, high cash gift and low cash gift group differences in measures of brain activity (electroencephalogram [EEG]). EEG activity will also be assessed in an in-home visit at age 1. To understand how family economic behavior, parenting, and parent stress and wellbeing change in response to income enhancement, we will assess treatment/control differences in family expenditures, food insecurity, housing and neighborhood quality; family routines and time use; parent stress, mental health and cognition; parenting practices; and child care arrangements at child age 2 and, for a subset of these measures, child age 1. This study will thus provide the first definitive understanding of the extent to which income plays a causal role in determining early child cognitive, socioemotional and brain development among low-income families. It will inform not only our basic and applied scientific understanding of early development, but also shedding light on the role of anti-poverty policies in promoting the wellbeing of poor children.

External Link(s)

Registration Citation

Citation

Duncan, Greg et al. 2019. "Baby's First Years." AEA RCT Registry. July 02. https://doi.org/10.1257/rct.3262-1.0

Former Citation

Duncan, Greg et al. 2019. "Baby's First Years." AEA RCT Registry. July 02. https://www.socialscienceregistry.org/trials/3262/history/49218

Sponsors & Partners

Interventions

Intervention(s)

Intervention:
$333 monthly cash gift payments to mothers each month for the first 40 months of their baby's lives via debit card.
Active Comparator:
$20 monthly cash gift payments each month for the first 40 months of their baby's lives via debit card.

Intervention Start Date

2018-05-09

Intervention End Date

2022-06-01

Primary Outcomes

Primary Outcomes (end points)

1. Child Language Development at age 36 months
2. Child Executive Function and Self-Regulation at age 36 months
3. Child Socioemotional Processing at age 36 months
4. Child IQ at age 36 months
5. Child Brain Function at age 36 months
6. Child Sleep at age 36 months
7. Child Body Mass Index at age 36 months
8. Child Health at 36 months
9. Child's School Achievement starting at age 6 years
10. Household Economic Hardship at ages 12, 24, 36 months
11. Maternal Physiological Stress at age 24 months
12. Parent-Child Interaction Quality at age 24 months

Primary Outcomes (explanation)

Here we list primary outcome measures. Details can be found in Appendix Tables 1 and 2 of the document "Analysis Plan and Measures".

1. Child Language Development at age 36 months: Language Processing measured using the Quick Interactive Language Screener- Language Processing Subscale (QUILS) (subject to change following pilot testing; we will estimate the statistical significance of the entire family of related measures in the Child Language Development outcome cluster using stepdown resampling methods for multiple testing)

2. Child Executive Function and Self-Regulation at age 36 months: Executive Function measured using one of the following (to be determined following pilot testing): Minnesota Executive Function Scale, EF Touch Executive Functioning, or Wechsler preschool and primary scale of intelligence—fourth edition- Working Memory Scale (WPPSI-IV) and Self-Regulation measured using Preschool Self-Regulation Assessment (we will estimate the statistical significance of the entire family of related measures in the Child Executive Function and Self-Regulation outcome cluster using stepdown resampling methods for multiple testing)

3. Child Socioemotional Processing at age 36 months: Social-Emotional Problems measured using the Brief Infant-Toddler Social and Emotional Assessment (BITSEA) and Behavior/Emotional Problems measured using the Child Behavior Checklist (we will estimate the statistical significance of the entire family of related measures in the Child Socioemotional Processing outcome cluster using stepdown resampling methods for multiple testing)

4. Child IQ at age 36 months measured using the Wechsler preschool and primary scale of intelligence—fourth edition (WPPSI-IV)

5. Child Brain Function at age 36 months: Resting brain function measuring Gamma, Alpha and Theta power using EEG resting high-frequency power (waves adjusted for multiple testing bias) (we will estimate the statistical significance of the entire family of related measures in the Child Brain Function outcome cluster using stepdown resampling methods for multiple testing)

6. Child Sleep at age 36 months using the Patient-Reported Outcomes Measurement Information System (PROMIS™) Sleep Disturbance (SD) Short Form

7. Child Body Mass Index at age 36 months measured using CDC scales

8. Child Health at 36 months measured using an additive index of six survey items (see Appendix Table 1 in the document titled "Analysis Plan and Outcome Measures" for items)

9. Child's School Achievement starting at age 6 measured using administrative test score data

10. Household Economic Hardship at ages 12, 24, 36 months measured using the household poverty rate using the Census Bureau's poverty thresholds by size of family (we will estimate the statistical significance of the entire family of related measures in the Household Economic Hardship outcome cluster using stepdown resampling methods for multiple testing)

11. Maternal Physiological Stress at age 24 months measured using maternal hair cortisol

12. Parent-Child Interaction Quality at age 24 months measured: Index of mother's positive parenting behaviors measured using the Parenting Interactions with Children: Checklist of Observations Linked to Outcomes (PICCOLO™) to code NICHD SECCYD Mother-Child Interaction Task (adapted script) (we will estimate the statistical significance of the entire family of related measures in the Parent-Child Interaction Quality outcome cluster using stepdown resampling methods for multiple testing)

Secondary Outcomes

Secondary Outcomes (end points)

1. Child Language Development at ages 12, 24, 36 months
2. Child Socioemotional Processing at ages 12, 24 months
3. Child Brain Function at age 12, 36 months
4. Child Physiological Stress at age 24 months
5. Child Sleep at ages 12, 24 months
6. Child Health at ages 12, 24 months
7. Child's Siblings' School Achievement starting at age 6 years
8. Child and Sibling School Behavior starting at age 6 years
9. Household Economic Stress at age 12, 24, 36 months
10. Social Services Receipt at ages 12, 24, 36 months
11. Mother's Labor Market and Education Participation at ages 12, 24, 36 months
12. Child-Focused Expenditures at age 12, 24, 36 months
13. Housing and Neighborhood Quality at ages 12, 24, 36 months
14. Family and Maternal Perceived Stress at ages 12, 24, 36 months
15. Maternal Happiness and Optimism at ages 12, 24, 36 months
16. Maternal Physiological Stress at age 12 months
17. Maternal Mental Resources at age 24 months
18. Maternal Mental Health at ages 12, 24, 36 months
19. Maternal Substance Abuse at ages 12, 24, 36 months
20. Chaos in the Home at ages 12, 24, 36 months
21. Maternal Relationship Quality at ages 12, 24, 36 months
22. Parent-Child Interaction Quality at ages 12, 24 months
23. Maternal Epigenetic Age at age 24 months
24. Maternal DNA Methylation at age 24 months
25. Frequency of Parent-Child Activity at ages 12, 24 months
26. Maternal Discipline at ages 12, 24, 36 months

Secondary Outcomes (explanation)

Here we list secondary outcome measures. Details can be found in Appendix Tables 1 and 2 of the document "Analysis Plan and Measures".

1. Child Language Development at ages 12, 24, 36 months: Language Milestones measured using Ages and Stages Questionnaire (ASQ)- Communication Subscale; Child Vocalizations measured using LENA Technology software (at age 24 months); Communicative Development measured using the MacArthur Communicative Development Inventories (at age 24 months); Verbal Comprehension measured using the Wechsler Preschool and Primary Scale of Intelligence-fourth edition (WPPSI-IV)- Vocabulary Subscale (at age 36 months) (we will estimate the statistical significance of the entire family of related measures in the Child Language Development outcome cluster using stepdown resampling methods for multiple testing)

2. Child Socioemotional Processing at ages 12, 24 months: Social-Emotional Problems measured using the Brief Infant-Toddler Social and Emotional Assessment (BITSEA)- Problem Scale; Social-Emotional Behavior measured using the NICHD SECCYD Mother-Child Interaction Task (positive/negative mood, activity level, sustained attention, positive engagement); Behavior/Emotional Problems measured using the Child Behavior Checklist (at age 24 months) (we will estimate the statistical significance of the entire family of related measures in the Child Socioemotional Processing outcome cluster using stepdown resampling methods for multiple testing)

3. Child Brain Function at ages 12, 36 months: Resting brain function measuring Gamma, Alpha and Theta power using EEG resting high-frequency power (waves adjusted for multiple testing bias) at 12 months; Language-related brain function measuring Alpha, Gamma and Theta power using EEG (waves adjusted for multiple testing bias) at 36 months (we will estimate the statistical significance of the entire family of related measures in the Child Brain Function outcome cluster using stepdown resampling methods for multiple testing)

4. Child Physiological Stress at age 24 months measured using child's Hair cortisol

5. Child Sleep at ages 12, 24 months measured using the Patient-Reported Outcomes Measurement Information System (PROMIS™) Sleep Disturbance (SD) Short Form

6. Child Health at ages 12, 24 months measured using an index of six items (see Appendix Table 1 in the document titled "Analysis Plan and Outcome Measures" for items)

7. Child's Siblings' School Achievement starting at age 6 years using administrative test score data

8. Child and Sibling School Behavior starting at age 6 years using administrative data

9. Household Economic Stress at ages 12, 24, 36 months: Food Insecurity measured using the U.S. Household Food Security Survey Module: Six-Item Short Form; Index of economic stress using an additive index of survey items, with higher score indicating higher stress (see Appendix Table 2 in the document titled "Analysis Plan and Outcome Measures" for items); Household Poverty measured using Census Bureau's thresholds (we will estimate the statistical significance of the entire family of related measures in the Household Economic Stress outcome cluster using stepdown resampling methods for multiple testing)

10. Social Services Receipt at ages 12, 24, 36 months measured by the number of benefits received by mother (see Appendix Table 2 in the document titled "Analysis Plan and Outcome Measures" for items)

11. Mother's Labor Market and Education Participation at ages 12, 24, 36 months: Time to labor market reentry from birth; Time to full-time labor market reentry from birth, Dichotomous indicator of mother's education and training activity participation (see Appendix Table 2 in the document titled "Analysis Plan and Outcome Measures" for items) (we will estimate the statistical significance of the entire family of related measures in the Mother's Labor Market and Education Participation outcome cluster using stepdown resampling methods for multiple testing)

12. Child-Focused Expenditures at age 12, 24, 36 months: Index of child-focused expenditures since birth; Total dollar amount of child-focused expenditures in the past 30 days; Cost of paid child care in dollars; Use of center-based care (see Appendix Table 2 in the document titled "Analysis Plan and Outcome Measures" for items) (we will estimate the statistical significance of the entire family of related measures in the Child-Focused Expenditures outcome cluster using stepdown resampling methods for multiple testing)

13. Housing and Neighborhood Quality at ages 12, 24, 36 months: Additive index of perceptions of neighborhood safety with higher score indicating feeling more safe; Additive index of housing quality; Additive index of items indicating experiences with homelessness; Indicator of three or more residential moves; Neighborhood Poverty Rate using census data (see Appendix Table 2 in the document titled "Analysis Plan and Outcome Measures" for items) (we will estimate the statistical significance of the entire family of related measures in the Housing and Neighborhood Quality outcome cluster using stepdown resampling methods for multiple testing)

14. Family and Maternal Perceived Stress at ages 12, 24, 36 months: Perceived Stress Scale (PSS) and Aggravation in Parenting Scale (we will estimate the statistical significance of the entire family of related measures in the Family and Maternal Perceived Stress outcome cluster using stepdown resampling methods for multiple testing)

15. Maternal Happiness and Optimism at ages 12, 24, 36 months: Global happiness item and HOPE Scale (we will estimate the statistical significance of the entire family of related measures in the Family and Maternal Happiness and Optimism outcome cluster using stepdown resampling methods for multiple testing)

16. Maternal Physiological Stress at age 12 months measured using maternal hair cortisol

17. Maternal Mental Resources at age 24 months measured using the Flanker Inhibitory Control and Attention Test

18. Maternal Mental Health at ages 12, 24, 36 months: PHQ-8 and Beck Anxiety Inventory (we will estimate the statistical significance of the entire family of related measures in the Maternal Mental Health outcome cluster using stepdown resampling methods for multiple testing)

19. Maternal Substance Abuse at ages 12, 24, 36 months: Index of frequency of alcohol and cigarette use; frequency of opioid use (see Appendix Table 2 in the document titled "Analysis Plan and Outcome Measures" for items) (we will estimate the statistical significance of the entire family of related measures in the Maternal Substance Abuse outcome cluster using stepdown resampling methods for multiple testing)

20. Chaos in the Home at ages 12, 24, 36 months: Home Environment Chaos Scale

21. Maternal Relationship Quality at ages 12, 24, 36 months: Index of frequency of arguing; Presence of physical abuse, Index of relationship quality (see Appendix Table 2 in the document titled "Analysis Plan and Outcome Measures" for items) (we will estimate the statistical significance of the entire family of related measures in the Maternal Relationship Quality outcome cluster using stepdown resampling methods for multiple testing)

22. Parent-Child Interaction Quality at ages 12, 24 months: Adult word count measured using LENA Technology software; Conversational turns measured using LENA Technology software (we will estimate the statistical significance of the entire family of related measures in the Parent-Child Interaction Quality outcome cluster using stepdown resampling methods for multiple testing)

23. Maternal Epigenetic Age at age 24 months measured using the Horvath Method

24. Maternal DNA Methylation at age 24 months

25. Frequency of Parent-Child Activity at age 12, 24 months measured by self-report of frequency of activities (see Appendix Table 2 in the document titled "Analysis Plan and Outcome Measures" for items)

26. Maternal Discipline at ages 12, 24, 36 months: indicator of use of spanking as a discipline strategy (see Appendix Table 2 in the document titled "Analysis Plan and Outcome Measures" for items)

Experimental Design

There are two arms of interventions. The experimental group or high cash gift group (compromising 40% of all participants) receives a monthly $333 cash gift for 40 months via debit card. The comparison group or low cash gift group (compromising 60% of all participants) receives a monthly $20 cash gift for 40 months via debit card. Interviewers are aware of experimental status at enrollment in order to activate the debit card after the participant consents to the study.

Experimental Design Details

Not available

Randomization Method

Computer-randomized and computer-accessed lists of experimental and control status for participants in each for the four sites. Interviewers are notified of experimental/control data during their enrollment interviews. All of the four metropolitan sites will have a 40/60 distribution of experimental and control cases.

Randomization Unit

Individual

Was the treatment clustered?

Yes

Experiment Characteristics

Sample size: planned number of clusters

4 metropolitan areas

Sample size: planned number of observations

1,000 mother/infant pairs, recruited in hospitals shortly after birth. Inclusion conditions (all must be met) include: 1. mother 18 years or older 2. household income below the federal poverty threshold in the calendar year prior to the interview, counting the newborn 3. infant admitted to the newborn nursery and not requiring admittance to the intensive care unit 4. residence in the state of recruitment 5. mother reports not "highly likely" to move to a different state or country in the next 12 months 6. infant to be discharged in the custody of the mother 7. Mother English or Spanish speaking (necessary for administration of instruments used to measure some of the child outcomes)

Sample size (or number of clusters) by treatment arms

400 in experimental group; 600 in control group. In New York City, we recruited 289 mother-infant dyads. We recruited 295 mother-infant dyads in the Omaha metropolitan area and the New Orleans metropolitan area. Lastly, we recruited 121 mother-infant dyads in the Twin Cities, MN.

Minimum detectable effect size for main outcomes (accounting for sample design and clustering)

.218 sd. The compensation difference between families in the experimental and control groups amounts to $313 per month and $12,520 over the course of the 40 months. This amount is in the range of income increases associated with child impacts of around .20 sd in studies of welfare experiments and the EITC. After accounting for likely 20% attrition, and in the absence of adjustments for sample clustering within hospitals or increased precision owing to the inclusion of baseline covariates in our impact estimates, the sample size of 800 at age 3, evenly divided between experimental and control groups, provides 80% statistical power to detect a .218 sd impact at p <.05 in a two-tailed test on cognitive functioning and family processes. Given the directional nature of all of our impact hypotheses, it could be argued that a one-tailed test is more appropriate, in which case our 80% power minimum detectable effect (MDE) size drops to .194 sd. The use of baseline covariates in estimation models will improve this power, while the use of bootstrap standard errors will decrease it, yielding offsetting effects of unknown but likely modest magnitudes.

Supporting Documents and Materials

There is information in this trial unavailable to the public. Use the button below to request access.

Request Information

IRB