99518 leans Democratic by roughly 22 points: about 61% of voters vote Democratic and 39% Republican. These figures are model estimates: Alaska did not have precinct-level voting records available for training, so the numbers above come from demographic and health features rather than local ground truth.
About 64% of adults in 99518 typically vote, near the U.S. average of about 62%. Among adults in 99518, ~39% vote Democratic, ~25% Republican, and ~36% don't vote. The map below shows estimated turnout by block group.
How 99518 compares
Among zip codes within 15 miles, 99518 leans more Democratic than 7 of 11 neighbors.
99518 runs about 35 points more Democratic than Alaska as a whole. Alaska leans Republican overall, while 99518 is one of the few Democratic-leaning pockets.
Politics vary noticeably by block within 99518. The northeast side is the most Democratic-leaning (D+29) and the west side is the least Democratic-leaning (D+15), a spread of about 14 points.
Why 99518 leans the way it does
This analysis examined 14,881 data points per zip code to find what predicts political lean and turnout. The items below are a few correlations that stood out for 99518, not a ranked or complete list of what matters most.
99518 votes against the grain of Alaska. Alaska leans Republican overall, while 99518 runs about 35 points more Democratic. A high never-married share predicts Democratic voting, and about 41% of adults in 99518 have never been married, above 88% of zip codes.
Population density and Democratic lean
Places with high population density tend to lean Democratic; 99518, AK sits in the top quarter nationally on this measure.
Why turnout in 99518 looks the way it does
Areas with high high-school completion turn out at higher rates. About 96% of adults in 99518 have completed high school, about 6 points above the Alaska average of 89%. Learn more about the findings and methodology on the political spectrum map.
Nearby Zip Codes
Zip Codes with Similar Populations
Sources and methodology
Precinct-level voting records used to fit the model come from Alaska Division of Elections, distributed by the Voting and Election Science Team. Demographic inputs come from the U.S. Census Bureau (ACS 5-year estimates and the 2020 Decennial Census). Health and environmental inputs come from the CDC (PLACES and the Environmental Justice Index). Land cover comes from the USGS and EPA. Election-day and lead-up weather come from PRISM 4km daily grids and the NOAA Global Historical Climatology Network. Mail-voting and election-administration patterns come from the MIT Election Lab's Survey of the Performance of American Elections. Block-group crime detail comes from CrimeGrade. Internet data and modeling support provided by ISPreports.org.
Modeling and analysis by the BestNeighborhood data science team. AK did not have precinct-level voting records available for training, so the figures here come from extrapolation across demographic, health, and land-use features rather than local ground truth. Full methodology and findings: political spectrum map.
Methodology reviewed by the BestNeighborhood data team. Last updated May 2026.