When the Berkeley City Council tackles citywide redistricting at tonight’s public hearing, it will be stuck with a population count from the 2000 census that, by most accounts, missed thousands of Berkeley residents.
The council must redraw district boundaries every 10 years in order to keep the population of each district roughly equal. The City Charter requires the council to base the whole process on numbers from the most recent census.
So now the city must rework its district map using numbers it is nearly certain are way off.
“My best estimate is that the census undercounted about 6,000 people,” said Patrick DeTemple, a senior systems analyst for the city who has been working with the Census Bureau. “At least three-fourths of them were in the area south of campus, so the vast majority were students.”
The presumed undercount disproportionately affected a swath that runs across the northern portions of districts 7 and 8, containing many of the university’s large dormitories and student apartment housing.
“This issue is generally what is fueling the entire redistricting process,” said Kriss Worthington, council member from District 7. “If students had not been undercounted, there would be incremental changes here and there, but there would not be the dramatic shifts in districts 7 and 8.”
And while redistricting may make the populations in each district equal according to the census numbers, an undercount would mean districts 7 and 8 will have several thousand more people than the other districts.
“The fluke of the student undercount means that districts 7 and 8 will actually have more residents that any other district in the city,” Worthington said. “So in those two districts, each individual person’s vote is worth a tiny bit less.”
The presumed undercount also means that any federal money doled out to cities on a per-resident basis would short Berkeley by 3-5 percent, depending on the actual size of the undercount.
University officials who worked closely with the Census Bureau in the months leading up to the count said the problems began when the bureau was supposed to put surveys in the students’ mailboxes.
“The census folks didn’t have the right numbers of envelopes,” said Je Nell Padilla of the university’s residential and family living/new student services department. “In some cases they had envelopes with no surveys in them. They missed whole chunks of people; they’d miss a building or whatever.”
There was supposed to be a second round of counting, Padilla said, when census workers would return to catch the students they had missed. But it never materialized.
“I went back and asked what they were doing with round two,” said Padilla. “And they said they didn’t have the staff to do it – even though it was part of the plan we had cooked up together. It was just crazy and disorganized.”
The city first suspected a massive undercount just days after the Census Bureau released its initial figures in May. According to the 2000 census, Berkeley has a population of 102,743. Its population in 1990 was 102,724.
“We saw that it was virtually the same as 10 years ago, even though we’ve had a big increase in housing,” said DeTemple. “Then we did a block-by-block comparison between 1990 and 2000. Specific blocks that contained dorms revealed radical drop-offs, when we knew that the dorms were still there and still inhabited.”
One block, said DeTemple, dropped from 1,070 residents in 1990 to exactly one resident in 2000.
“That was when we knew there was a very specific and serious problem.”
The city alerted the Census Bureau of the problem within weeks. Soon after, a report from California’s department of finance confirmed the city’s findings, estimating that 6,000-7,000 Berkeley residents were missing from the census tallies.
It is the bureau’s policy not to recount. The bureau is moving the city’s complaint through its count question resolution program, which can adjust the totals if residents were counted but missed in the final tally because of processing errors.
Still, according to the program guidelines, any Berkeley residents who were never counted to begin with cannot be added at this point. Moreover, even if the bureau adjusts its official numbers at all, it will likely be too late to affect redistricting.
That leaves the city few options to correct the problem.
“We haven’t found a way,” said Worthington. “There is a possibility that after we get new numbers we could redistrict again and get people all bent out of shape again. But I don’t know if people will want to go through it all again.”