Racial Disparities

Overview

Racial disparities in New York City’s child welfare system have been well-documented. Citywide, almost 45% of Black and Latino children experience an investigation of their family by age 18 compared to 19% for White children. In a city where about 60% of children are Black and Latino, they account for about 90% of children in the foster system. Black families also experience the harshest impacts. Black children are 10 times as likely to enter foster care than white or Asian children and twice as likely as Latino children.

Yet public data masks the true scope of child welfare’s impact on Black and Latino families—and especially Black families—in many NYC neighborhoods.

This brief examines rates of child welfare investigations and foster care entries for Black, white, Latino and Asian children in NYC using zip code-level data from 2019. It also examines these racial disparities through the lens of child poverty.

Investigation rates of Black children by zip code make clear that:

  • Black children are extraordinarily vulnerable to investigations no matter how rich or poor the neighborhood they live in.
  • Black families in 68% of NYC zip codes live with investigation rates above the national average.
  • While only 8 zip codes in the city had overall investigation rates of 1 in 10 children or higher in a 2019, 50 zip codes had these staggering investigation rates for Black children. That’s almost one-third of all zip codes.

Latino children also face pervasively high rates of investigations, with investigation rates above the national average in 45% of zip codes, and enter foster care at five times the rates of White and Asian children.

Unlike Black children, Latino, White and Asian children all show lower investigation rates in neighborhoods where child poverty is lower. Neighborhood child poverty rates do not appear to have the same protective effect for Black children. In fact, Black children face extremely high investigation rates in dozens of well-off and majority white neighborhoods, such as Brooklyn Heights or Boerum Hill.

These data suggest that legislative and policy remedies in NYC cannot be race-blind. As Dr. Tricia Stephens has written about Black friends living in these “upgraded segregation settings” and facing investigations: “My friends cannot protect themselves from the scrutiny that is anti-Black racism…these Black mothers fall solidly within the crosshairs of regulation and social control.”

The disproportionate representation of Black families within the child welfare system has long been researched ahistorically – examined from the perspective of intrafamilial stressors/vulnerabilities/fragility without accounting for the structural racism and social disinvestment that impact Black families’ ability to achieve and maintain stability, as Dr. Stephens has written.

In reports from Rise and the Narrowing the Front Door Work Group, NYC parents impacted by family policing and their allies have advanced legislative and policy visions to constrain the scope and power of the child welfare system, build community care and healing, and make a broad reparative investment in Black families and communities. This brief offers information on where efforts to address the conditions driving child welfare involvement can have the most impact.

How to Use This Brief

  • Contact info@familypolicynyc.org with any questions. Our goal is to make sure parents and youth impacted by the child welfare system, advocates, community organizations, policymakers, media and others have access to information about the child welfare system’s impact in New York City.
  • For more information on hotline calls and investigations, see FPP's brief on hotline calls. You also can learn more here about how to understand ACS' publicly available data about hotline calls and investigations.

 

 

Data analysis by Cat Pisciotta, data modeling by Dr. Frank Edwards and brief text by Nora McCarthy. Data review by Ryan Brown. All child welfare data provided by NYC's Administration for Children's Services (ACS).

Black families face extraordinary overexposure to child welfare involvement

Black and Latino children are over-represented in child welfare involvement while white and Asian children are under-represented.

Download: Chart Image, Table Image, Data

 

Note: This brief uses Census child race and ethnicity data for analysis and therefore refers to children, not families, in most data. Percent of children in investigations and percent of children entering foster care do not sum to 100% here because “other” race/ethnicity is not included here. Please see methodological notes at the bottom of the page regarding child and resident (all ages) populations. Prevalence rates reflect aggregates, not averages. “Foster care entries” includes ACS removals as well as voluntary placements, PINS, abandonment, or other placements.

 

Race Ethnicity Chart

For Black children, 50 zip codes had investigation rates of 1 in 10 children or higher

Only 19 zip codes show similar rates for Latino children, 7 zip codes for white children and 5 zip codes for Asian children.

The investigation rate for Black children is above the national average of 47.2 per 1,000 children in 111 out of 163 zip codes, which is 68% or two-thirds of NYC zip codes.  The citywide investigation rate in NYC is 42 investigations per 1,000 children, which is slightly below the national average.

Download: Image

 

 

Note: Only about 170 zip codes (somewhere between 160 and 180 depending on the racial/ethnic group) had child census data available with which to calculate rates. Additionally, some zip codes were excluded if their census population was lower than the number of children of a given racial/ethnic group in investigations.

Investigation Rate: Black Children, 2019 Map

For Latino children, 19 zip codes had investigation rates of 1 in 10 children of higher

The investigation rate for Latino children is above the national average of 47.2 per 1,000 children in 79 zip codes (out of 176 zip codes with data), accounting for 45%, or almost half of neighborhoods.

Download: Image

Investigation Rate: Latino Children, 2019 Map

Only 1% of white or Asian children live in zip codes with investigation rates above the national average

The white investigation rate is above the national average of 47.2 per 1,000 children in 16 zip codes out of 177 with data. That is almost a tenth of neighborhoods but less than 1% of white children in NYC live in zip codes with these investigation rates.

In 19 zip codes out of 173 with data, the Asian/PI investigation rate is above the national average (that’s 11% of neighborhoods) but only about 1.1% of Asian/PI children live in zip codes with these investigation rates.

Investigation rates of 1 in 10 children or higher were found in in 7 zip codes for white children and 5 zip codes for Asian children, compared to 50 zip codes for Black children and 19 for Latino children.

Download: Map Image 1, Map Image 2

 

Investigation Rate: White Children, 2019 Map

 

Investigation Rate: Asian Children, 2019 Map

More than half of all children in investigations live in majority Black and Latino neighborhoods

While only 35% of all NYC children lived in majority Black/Latino zip codes, 51% of all children in investigations lived in these neighborhoods.

Meanwhile, 25% of all NYC children lived in majority white/Asian zip codes but only 11% of children in investigations lived in these neighborhoods.

Citywide, over half of children involved in investigations came from just 31 zip codes. Nearly all of those (87%, or 27 zip codes) are zip codes where Black or Latino children (or both) are heavily concentrated.

Download: Image, Data

 

 

Note: Majority Black/Latino neighborhoods are 75% or more Black/Latino residents; n=44. Majority White/Asian neighborhoods are 75% or more White/Asian residents; n=61. Per the American Community Survey 5 year estimates (2015-2019), there were 1,753,499 children in NYC, 35% of which (608,245) lived in majority Black/Latino neighborhoods and 25% of which (433,003) lived in majority White/Asian neighborhoods. Total children involved in investigations (2019) = 78,041. From majority Black/Latino neighborhoods: 40,018. From majority White/Asian neighborhoods: 8,282.

Child welfare impacts are highly concentrated and segregated

Citywide, Black, white, Latino and Asian children tend to live in a relatively small number of distinct zip codes (about 20-27 zip codes for each group), and child welfare impacts for each group are also highly concentrated. 

For each of these groups, about 25% of investigations take place in just 9 zip codes (10 for white children). Efforts to address community conditions driving child welfare involvement would have the greatest impact in these neighborhoods. 

About 50% of children for each group live in just 25 zip codes, and 50% of all investigations for each racial/ethnic group also take place in these zip codes.

Foster care entries citywide are even more concentrated than investigations, with half of all foster care entries for Black, Latino and white children concentrated in about 22 zip codes for each group.

Download: Image 1 and Data 1, Image 2 and Data 2, Image 3 and Data 3, Image 4 and Data 4

 

 

Note: For Black, Latino and Asian children, 75% of investigations are concentrated in about 50 zip codes out of 180 with census data. A final 25% of child welfare involvement is spread across 125-130 zip codes for these groups. White children’s child welfare involvement is slightly less concentrated, with 67 zip codes accounting for 75% of investigations and 111 zip codes accounting for the final quarter.

For Asian children, foster care entry is even more concentrated. Just 12 zip codes account for half of all foster care entries for Asian children. For Black, Latino and white children, 75% of foster care entries took place in about 45 zip codes, while 75% of foster care entries of Asian children took place in 21 zip codes.

 

Citywide child poverty rates do not correlate with child welfare involvement rates by race

If poverty were the only factor, investigation and foster care rates by race and ethnicity would mirror citywide averages for child poverty, but they don’t.

Child poverty rates would suggest that Black and Latino children experience investigations and removals at approximately twice the rate of white and Asian children.

However, Black children were 6 times more likely to experience an investigation than white children and 10 times more like to be separated from their families than white and Asian children. Latino children were 5 times more likely to experience an investigation than white children and 5 times more like to be separated from their families than white and Asian children.

Download: Image, Data

 

Note: Percent of children in poverty reflects the NYC 3-year average, 2019. “Other” race/ethnicity is excluded from presentation here. Prevalence rates reflect aggregates, not averages. Black children were 4 times more likely to experience an investigation than Asian children, and Latino children were 3 times more likely.

Black children are extraordinarily vulnerable to investigations no matter how rich or poor the neighborhood they live in

This chart uses modeling to represent a trend line for the city, evaluating the relationship between exposure to ACS investigations and the child poverty rate for four racial and ethnic groups. Instead of showing every zip code, the model summarizes the overall trend. This regression model allows us to see the average relationship between race-specific neighborhood poverty rates and ACS investigations across different kinds of NYC neighborhoods.

Importantly, average levels of ACS investigations of Black children annually are more than 100 per 1,000 in both low and high Black child poverty neighborhoods. In other words, we find that around 10% of Black children can expect an ACS investigation annually, regardless of the poverty levels of Black children in their neighborhood. These rates are higher than average investigation rates for all other groups of children.

Latino children also are, on average, investigated at much higher rates than white children, regardless of child poverty rates.

In high white poverty neighborhoods, we observe much lower levels of ACS investigations for white children than for Latino or Black children in neighborhoods that have comparably high Black or Latino child poverty.

Download: Image

 

Note: The trend lines shown illustrate our model’s expectation for ACS investigation rates in NYC zip codes with different levels of child poverty across four racial and ethnic groups. Black, Latino and white children are exposed to higher levels of ACS investigations in zip codes with higher levels of child poverty. We find no clear relationship between AAPI child poverty rates and AAPI investigation rates, which are, on average, similar to white investigation rates.

Race Ethnicity Child Poverty Rate Chart

Black children are most exposed to investigations in predominantly white and Asian neighborhoods

In this chart, we estimate a second model that considers how the racial composition of a neighborhood and its child poverty rate relate to the rates at which children of different groups are investigated by ACS. Unlike the previous chart, which used race-specific poverty rates, this model uses zip code-level poverty rates for ALL children.

We then estimate expected levels of ACS investigations for each group of children in four typical NYC neighborhood types: 1) Low child poverty, predominantly white or AAPI, like the Upper West Side; 2) Low child poverty, predominantly Black or Latino, like Jamaica, Queens; 3) High child poverty, significantly white or AAPI, like Williamsburg, Brooklyn; 4) High child poverty, predominantly Black or Latino, like East Harlem.

We show that Black children are investigated at extraordinarily high rates in zip codes that are predominantly white neighborhoods with low overall child poverty. In these zip codes, about 15% of Black children can expect an ACS report in a given year. In predominantly Black and Latino neighborhoods with low child poverty, by contrast, Black children are investigated at an average rate of about 6% annually. In low-poverty, mostly white neighborhoods, Black children are investigated 2.5 times more frequently to ACS than are Black children in low-poverty mostly nonwhite neighborhoods.

We also show that Black children are reported at relatively lower rates in predominantly Black and Latino zip codes with high child poverty than they are in high child poverty neighborhoods that are predominantly White and AAPI. The racial composition of neighborhoods has a powerful relationship to ACS investigation rates across poverty contexts.

Download: Image

Race Ethnicity ACS Investigations Chart

Black children in a small number of high-income white and Asian zip codes have extraordinarily high investigation rates

34 zip codes with high Black or Latino poverty and low white or Asian poverty include only 5.6% of Black children citywide but account for 10% of all investigations of Black children. Many of these zip codes have public housing complexes, family shelters, or both. Latino children in these neighborhoods do not have similarly high investigation rates. Black children in these zip codes do not appear to benefit from overall low-poverty neighborhood conditions.

Download: Image, Data

 

Note: Average investigation rate of 163 for Black children and average investigation rate of 67 for Latino children in these zip codes. The percent of Black/Latino residents in these zips (n=34) was an average of 28% (minimum of 0 to maximum of 59%), compared to the citywide average of 46%. Investigations of Black children n=3,113; 25,025 Black children. Investigations of Latino children n=3,659; 67,087 Latino children. Investigations of Asian/PI children n=748; 51,920 Asian/PI children. Investigations of white children n=1,623; 112,023 white children. Percent of children in poverty citywide average is 20%; average in these zips is 15%.

What are the investigation numbers by race in my zip code?

You can use this table to see the number and percent of investigations by race in every zip code. Data is by numbers of children, not families, except for the number of investigations.

To explore: Click on the each label (such as “zip code” or “white”) to sort the sheet by that data type from the lowest number to the highest, and click again to show highest to lowest.

 

Note: A cell value of ** indicates that there were 5 or fewer investigations in the zip code 2019. Those data have been blinded to protect confidentiality, at ACS’ request. Total children in investigations and numbers for Black, Latino, white and Asian children also are blinded in zip codes with 6-10 investigations.

For context, please include the number of total children in INV/entering FC if using the percentages provided.

Child Census data comes from the American Community Survey’s 5-year (2015-2019) estimates. In NYC Administration for Children’s Services (ACS) data, Black=Black non-Hispanic/Latino, Latino=Hispanic/Latino (any race), White=White non-Hispanic/Latino, and Asian/PI=Asian/PI non-Hispanic/Latino.
In the American Community Survey Census data, Black=Black (whether Hispanic/Latino or not), Latino=Hispanic/Latino (any race), White=White non-Hispanic/Latino, and Asian/PI=Asian/PI (whether Hispanic/Latino or not), due to the data available by zip.

No rates are provided in any case where the child census for the zip code is 0 (meaning either there is no data for that zip code in the census or the child population is actually 0). No rates are provided in any case where the child census for the zip code is larger than the number of children in investigations.

ACS removes Black children in an investigation almost twice as often as Latino or white children

Asian children are the least likely to be separated from their families in an investigation.

Download: Image, Data

 

Note: “Foster care entries” includes ACS removals as well as voluntary placements, PINS, abandonment, or other placements.

More than half of all children who enter foster care live in majority Black and Latino neighborhoods

While only 35% of all NYC children lived in majority Black/Latino zip codes, 55% of all children separated from their families lived in these neighborhoods.

Meanwhile, 25% of all NYC children lived in majority white/Asian zip codes but 7% of children separated from their families lived in these neighborhoods.

Citywide, over half of children entering foster care came from 27 zip codes. Many of those (81%, or 22 zip codes) are zip codes where Black or Latino children (or both) are heavily concentrated.

Download: Image, Data

 

Note: Majority Black/Latino neighborhoods are defined as 75% or more Black/Latino residents; n=44. Majority White/Asian neighborhoods are 75% or more White/Asian residents; n=61. Per the American Community Survey 5 year estimates (2015-2019), there were 1,753,499 children in NYC, 35% of which (608,245) lived in majority Black/Latino neighborhoods and 25% of which (433,003) lived in majority White/Asian neighborhoods. Total children entering foster care (2019) = 3,430. From majority Black/Latino neighborhoods: 1,873. From majority White/Asian neighborhoods: 247.

Almost 40 percent of Black children citywide lived in zip codes where the percent of families separated after an investigation was well above the citywide average

Download: Image, Data

 

Note: The citywide average was 3.7 (aggregate value: 4.4). This chart presents zip codes with values 5.8 or more, which was the mean/aggregate for Black children. Only zip codes with 1,000 or more Black children in the 2019 census are presented here for brevity. 23 other zip codes also had similarly high percentages (for a total of 56 zip codes).

10% of Latino children citywide lived in zip codes where the percent of families separated after an investigation was well above the citywide average

Percentages for Latino children typically hovered around the citywide numbers.

Download: Image, Data

 

Note: The citywide average was 3.7 (aggregate value: 4.4). The Latino average was 3.7 (aggregate value: 3.5). This chart presents zip codes with values 5.8 or more, which was the mean/aggregate for Black children (the highest value among all racial/ethnic groups). Only zip codes with 1,500 or more Latino children in the 2019 census are presented here for brevity. 19 other zip codes also had similarly high percentages (for a total of 36 zip codes).

What are the foster care entry numbers by race in my zip code?

You can use this table to see the number and percent of foster care entries by race in every zip code. Data is by numbers of children, not families.

To explore: Click on the each label (such as “zip code” or “white”) to sort the sheet by that data type from the lowest number to the highest, and click again to show highest to lowest.

Note: A cell value of ** indicates that there were 5 or fewer families with children entering foster care in the zip code 2019. Those data have been blinded to protect confidentiality, at ACS’ request. Total children entering foster care and numbers for Black, Latino, white and Asian children also are blinded in zip codes with 6-10 investigations.

For context, please include the number of total children in INV/entering FC if using the percentages provided.

Child Census data comes from the American Community Survey’s 5-year (2015-2019) estimates. In NYC Administration for Children’s Services (ACS) data, Black=Black non-Hispanic/Latino, Latino=Hispanic/Latino (any race), White=White non-Hispanic/Latino, and Asian/PI=Asian/PI non-Hispanic/Latino.
In the American Community Survey Census data, Black=Black (whether Hispanic/Latino or not), Latino=Hispanic/Latino (any race), White=White non-Hispanic/Latino, and Asian/PI=Asian/PI (whether Hispanic/Latino or not), due to the data available by zip.

No rates are provided in any case where the child census for the zip code is 0 (meaning either there is no data for that zip code in the census or the child population is actually 0). No rates are provided in any case where the child census for the zip code is larger than the number of children in investigations.

Methodological Notes

Census child race and ethnicity data comes from the 2015-2019 American Community Survey 5 year estimates. In the available data, white children are white non-Hispanic/Latino and Latino children are Hispanic/Latino (any race). Black children include both Latino and non-Latino children in the census data, as do Asian/PI children.

Census population (all age residents) data comes from the 2015-2019 American Community Survey 5 year estimates. Data include Black non-Latino, Latino (any race), White non-Latino, and Asian/PI non-Latino.

NYC Administration for Children’s Services data includes Black non-Latino, Latino (any race), white non-Latino, and Asian/PI non-Latino. Other race/ethnicity is not presented in the analyses here.

When calculating rates, zip codes are excluded if they do not have census data or if the total census (overall or for a given R/E group as applicable) is greater than the value in the CW stage of interest. For aggregate calculations, no zip codes were excluded.