Beijing's license plate lottery policy was originally designed to curb the growth of local vehicle population. However, the avoidance behaviors such as local residents registering their cars in neighboring cities offsets the policy effect. Using the city-level data collected from the CEIC China Premium Database, this study quantitatively identifies the causal effect of the implementation of Beijing's car ownership restriction policy on the growth of private vehicles in neighboring cities. We first use a synthetic control method creating a weighted combination of non-treated cities for each treated city, then employ a difference in difference approach to test the policy leakage effect. The result shows a causal effect of 5.9% on average of Beijing's car ownership restriction policy on the growth of private vehicles in neighboring cities, which amounts to 549 thousand cars. The magnitude of the policy leakage declines by 7.1% every 100 km of driving distance away from Beijing within the 500 km boundary. Our result suggests that as much as 35.4% of the growth in private vehicle population that could have been reduced by the policy simply spilled over to neighboring cities. The significance of the policy leakage necessitates putting Beijing's congestion issues in a broader context and executing regional collaboration. Accordingly, we give some policy suggestions such as parking lots construction, transit system improvement and job- housing co-location.