Under-5 mortality rate
usable no number here
The design ran but cannot isolate an effect, so there is no number.
Not an estimate. Widest trend gap: 1999–2007 vs 2007–2010, +3.79 per 1,000 live births.



Causal Presidents · step 4 of 6
usable no number here
The design ran but cannot isolate an effect, so there is no number.
Not an estimate. Widest trend gap: 1999–2007 vs 2007–2010, +3.79 per 1,000 live births.
usable no number here
The design ran but cannot isolate an effect, so there is no number.
Not an estimate. Widest trend gap: 2007–2010 vs 2015–2023, -30.74 per 100,000 births.
usable with a warning about the money channel
The result is about the money channel, not this outcome.
Not an estimate. Widest trend gap: 2007–2010 vs 2010–2015, +11.19 % surviving infants.
usable with a warning about the money channel
The result is about the money channel, not this outcome.
What we have: 8 observed years, 1999 to 2022 (a trend needs 20).
the check it needed failed a declared gap, not a zero
There are too few observed years to estimate on.
What we have: 7 observed years, 2003 to 2024 (a trend needs 20).
the check it needed failed a declared gap, not a zero
There are too few observed years to estimate on.
What we have: 8 observed years, 2000 to 2010 (a trend needs 20).
the check it needed failed a declared gap, not a zero
There are too few observed years to estimate on.
What we have: 9 observed years, 1999 to 2010 (a trend needs 20).
the check it needed failed
+0.4678 95 % interval from -0.0833 to 1.0189
In the same state and month, a 1 % higher federal allocation received by the state goes with 0.47 % more reported security fatalities.
The interval includes zero.
Not an estimate. Widest trend gap: 2007–2010 vs 2010–2015, -209.99 in the count.
On 3,888 observations over 36 units, 2016-12-01 to 2025-12-01.
Verdict: estimated_not_identified.
The check that decided it: the lead-shift p-value is 0.167: 10 of 60 mis-timed doses return an estimate at least as large as the estimate itself.
the check it needed failed
+0.5278 95 % interval from 0.1467 to 0.9088
In the same state and month, a 1 % higher federal allocation received by the state goes with 0.53 % more reported security incidents.
The interval stays on one side of zero.
Not an estimate. Widest trend gap: 2007–2010 vs 2010–2015, -13.27 in the count.
On 3,888 observations over 36 units, 2016-12-01 to 2025-12-01.
Verdict: estimated_not_identified.
The check that decided it: the lead-shift p-value is 0.167: 10 of 60 mis-timed doses return an estimate at least as large as the estimate itself.
usable with a warning about the money channel
The result is about the money channel, not this outcome.
Not an estimate. Widest trend gap: 2015–2023 vs 2023–2026, -3.27 %.
usable with a warning about the money channel
The result is about the money channel, not this outcome.
Not an estimate. Widest trend gap: 2015–2023 vs 2023–2026, -5.55 %.
usable no number here
The design ran but cannot isolate an effect, so there is no number.
Not an estimate. Widest trend gap: 2015–2023 vs 2023–2026, +1.31 %.
Registry vintage 2026-11-01