A state implemented mandatory seatbelt enforcement on January 1, 2024. Highway fatality data from a sample of counties:

| | Before 2024 | After 2024 | |---|---|---| | Enforcement counties | 120 deaths | 105 deaths | | Non-enforcement counties | 95 deaths | 88 deaths |

A simple comparison in enforcement counties is $(105 - 120) = -15$ deaths. However, a difference-in-differences (DiD) analysis calculates:

$\text{DiD Effect} = [\text{After}_{\text{enforce}} - \text{Before}_{\text{enforce}}] - [\text{After}_{\text{no-enforce}} - \text{Before}_{\text{no-enforce}}]$

What is the DiD estimate of the policy effect?