The Hawaii Reapportionment Commission has reversed its decision to count non-permanent military and college students as residents for determining Hawaii’s political boundaries.

The commission on Monday voted 5-3 in favor of a model that would remove 16,458 active-duty military and out-of-state college students from the state’s population base. The so-called extraction would not affect the number of Hawaii House and Senate seats assigned to each island.

The group is working toward a Friday deadline to file final proposals with the Chief Elections Officer, who then has until Oct. 10 to publish the plans. Revised maps reflecting the new base population should be available online by Tuesday afternoon.

The commission had previously held that there was no accurate way to separate non-resident military and students from the general population based on where they live, as has been done in previous years. The commission voted 8-1 in June to include these two populations as well as sentenced felons — the basic Census population — as part of political district populations.

Last week the commission held two hearings on the Big Island, where members heard protests against counting military and threats of lawsuits if it insisted on counting non-resident military and students.

On Monday, reapportionment staff presented new data at the commission’s meeting showing that it would be possible to determine whether some military personnel were non-residents. They said that Census data did show military living in group quarters on military bases, rather than in the community, by Census tract.

Based on that new data, the staff came up with three scenarios to remove some military and students from the state’s population base.

The approved model — dubbed Extraction Model A — uses Census data that identifies military personnel who live in group quarters on military bases in Hawaii, but the figures do not distinguish between permanent and non-permanent residents within the group. That means Hawaii residents and registered voters who happen to live in these quarters would be extracted from the population for reapportionment purposes. Commission Chairwoman Victoria Marks, who supported the model, said that being a registered voter in Hawaii would be a strong indicator that the person was a permanent resident, but that there wasn’t enough time to try to weed those people out.

Reapportionment staff said the A model has the least chance of removing permanent residents who live on military bases. It would also remove out-of-state students whose full addresses were provided by their universities and who claim permanent residence in another state.

In addition to Marks, commission members Lorrie Lee Stone, Clarice Hashimoto, Calvert Chipchase and Terry Thomason voted in favor of the A model. Members Dylan Nonaka and Elizabeth Moore voted against any extraction. Commissioner Anthony Takitani was in favor of the other two models that would have extracted higher numbers of military and students. (Commission member Harold Masumoto was absent.)

Here’s a breakdown of Extraction Model A:

Island Unit Number Extracted Adjusted Population
Oahu -15,660 937,547
Hawaii -793 184,286
Maui -4 154,920
Kauai -1 67,090

Extraction Models B and C proposed extracting a total of 73,552 and 79,821 military and students from the population, respectively. The B model would include all of the A model, plus military members living on bases in non-group housing. The C model would include all of A and B, plus university students for whom only ZIP codes were provided versus addresses.

In voting against all three scenarios, commissioner Nonaka said: “There is a major problem I have — simply that it focuses on a certain status and class of a person and the geography of where they reside, and not whether they are permanent or non permanent residents … If we are going to exclude people based upon their nonresidency, we have to exclude them based on their nonresidency and not any other criteria or data.”

Commissioner Chipchase, who supported the A model, said he felt that counting military in group housing would be a safe route.

“We cannot determine which Census block where people who claim another state are from,” he said, referring to military personnel. “The next best thing is to look at military bases. Group housing is a less permanent form of housing and tends to be more transient.”

Commissioner Takitani, who was the only member who voted against counting military in the first place, said he favored B or C because they proposed taking more people out.

Follow Civil Beat on Facebook and Twitter. You can also sign up for Civil Beat’s free daily newsletter.