Permanency, prevention, and stability — three indicators that together describe whether a child
welfare system is keeping families intact, returning children to safe homes promptly, and
making those returns stick.
Most state report cards focus on a single number. This explorer lets you weight each measure
yourself — including a fourth dimension on placement stability — and watch the rankings
shift in response to what you think a healthy system looks like.
The Tile map gives every jurisdiction equal visual weight so small states aren't overlooked.
The Ranking view is the canonical sorted-bar chart. Relationships lets you plot
any two metrics against each other to spot trade-offs.
Combined performance · all jurisdictions
Click a jurisdiction to inspect. 45 of 52 scored
Lower
0255075100
HigherGrey tiles = held out (DC, NC, PA, WA, WV, WY)
#
Jurisdiction
Combined Score
Pts
Points are coloured by combined performance score. Hover to identify; click to lock in.
Try this
Turn on placement instability and look at
South Carolina. It has the best permanency record in the country — 53.3% within a year — and
the worst placement stability, at 12.29 moves per 1,000 days. The same system that moves
children out of care fastest also moves them around the most while they are in it. Whether
that reads as success depends entirely on which slider you believe.
Source · U.S. Administration for Children & Families · CFSR 4 Data Profile Dashboard ·
tableau-public.acf.gov
45 of 52 jurisdictions scored · 6 held out for suppressed data · Puerto Rico not shown
Methodology
How the combined score works
A state that exits children to permanency quickly may simply have moved them out of foster care
before families were ready — and may see them re-enter months later. Looking at a single
measure can mislead. This tool blends three federal indicators — and optionally a fourth — into
a tunable composite so you can see how the picture changes under different priorities.
1. The three base indicators
Entry rate (entries ÷ child population × 100,000) — fewer children
entering foster care, all else equal, is better. Permanency in 12 months — share of entrants who exit to reunification, adoption, or
guardianship within a year. Higher is better. Re-entry within 12 months — share of exits that return to care within a year. Lower
is better.
Both directions are contestable, and it is worth saying so plainly. A low entry rate can mean
strong prevention — or it can mean maltreatment is going undetected. Fast permanency can mean
children are going home promptly — or that they are going home before families are ready. The
arrows below encode one reading; the sliders exist so you can test others.
2. These are risk-standardized figures
The three child-welfare indicators are ACF's risk-standardized performance values, not
raw observed rates. ACF adjusts each state's result for the characteristics of the children it
serves, so states with different caseloads can be compared. Observed and risk-standardized
figures can differ substantially: Michigan's observed permanency is 17.6% against a
risk-standardized 18.9%.
3. The three measures come from three different periods
They are the most recent published window for each indicator, and those windows do not line up.
Permanency uses the cohort entering care in ACF period 23B24A; re-entry uses
24A24B; placement stability uses 25A25B. Each risk-standardized value
also pools two years of data to stabilise the estimate. A composite therefore describes a system
across roughly two and a half years, not a single snapshot.
4. Standardize each to 0–100
Each indicator is rescaled across the scored jurisdictions using min-max normalization so that
the best performer earns 100 and the worst earns 0. The direction of "better" is flipped for
entry and re-entry so that higher always means better performance. Because the ends of
the scale are set by whichever jurisdiction happens to be most extreme, a score is a statement
about rank within this group — not an absolute standard, and not comparable across a redraw of
the group.
5. Combine with weights you choose
Overall = w_E · EntryScore + w_P · PermScore + w_R · ReentryScore
Weights must sum to 1; the sliders auto-normalize as you move them. Equal weighting is a
reasonable default, but a focus on prevention might justify a larger entry-rate weight; a focus
on system stability might justify weighting re-entry more heavily. Placement instability — moves
per 1,000 days in care — can be toggled on as a fourth dimension.
Ranks are less precise than they look
Every indicator carries a confidence interval, shown beside each measure when you select a
jurisdiction. They are frequently wide enough to overlap: South Carolina leads on permanency at
53.3% and Colorado follows at 53.1%, but their intervals are 51.2%–55.3% and 51.0%–55.2% — very
nearly the same range, so first and second place on that measure are not meaningfully apart.
Treat neighbouring ranks as indistinguishable, and read the map for broad bands rather than
exact positions.
Who is missing, and why
ACF suppresses an indicator when a state's submitted data does not meet its quality standard.
A jurisdiction missing any indicator cannot sit on the same scale as one measured on all four,
so 45 of 52 jurisdictions are scored and six are shown on the map without a score:
Pennsylvania, Washington and Wyoming — permanency, re-entry and placement stability all
suppressed. West Virginia — permanency and re-entry suppressed; only placement stability (5.94) is
published. District of Columbia and North Carolina — re-entry suppressed, and no entry rate on a
basis comparable to the rest. Their other measures are published: permanency 18.9% and 28.1%,
placement stability 6.37 and 6.83. They will be added once a comparable entry rate is available.
Puerto Rico appears in the federal data with all indicators suppressed and is not shown.
Holding these out matters for reading the rest. West Virginia has by far the highest entry rate
in the country, and while it was included the entry-rate scale was anchored to it — compressing
every other jurisdiction into the top half of that measure. With it held out, the scale runs from
New Jersey to Montana and uses its full range.
Source: U.S. Administration for Children & Families, CFSR 4 Data Profile Dashboard
(Risk-Standardized Performance), retrieved 13 August 2026. Entry rates are computed separately
from foster care entry counts and child population.