Questions raised about population forecasts

The rationale for expanding the urban growth boundary comes from population forecasts generated by the City of Sisters Planning Department. These population forecasts have come under scrutiny by local watchdog groups and the math department at Central Oregon Community College.

The City of Sisters planning department's population forecast showed that the city had a shortage of 90 acres for residential housing. Proposed new "Neighborhood Centers" total 111 acres.

Last year, the Sisters Forest Planning Committee (SFPC) filed an appeal with the Land Use Board of Appeals concerning county-wide population forecasts. Deschutes County voluntarily remanded and voided its population forecast acknowledging that it has to be redone.

The SFPC contends that Sisters' population figures are artificially inflated by incorrect data.

The data used to determine the proposed urban growth boundary expansion says that by the year 2025 the Sisters population will reach 4,837. The population estimate for the year 2000 was 975 people. So the city planning department has predicted that in 2025 Sisters will have grown by five times the population of the year 2000.

This population forecast was generated by analyzing the number of building permits for residential units, issued after the sewer system went in during the year 2000.

Paul Dewey of the SFPC said, "No other city in Deschutes County is doing their analysis on the basis of future anticipated building permits. There's no factual basis for making up a bunch of numbers for 20 years based on how many building permits you think you're going to be issuing.

"It's too abstract," he said. "Sisters is still lacking the fundamental requirements of explaining their methodology and showing where they got their numbers."

City Planner Neil Thompson defended the city's methodology.

"This statistical model is what Portland State University uses," Thompson said. "They (PSU) are the experts."

The average number of building permits before the sewer system was 19 per year. After the sewer system, from 2001 to 2003, the average jumped up to 75 permits. The population forecast uses an 81 permit average to project the future for 2004 and beyond.

Thompson explained, "To generate this data we also used what's called 'regression to the mean.'"

This concept extrapolates from past events to come up with an average for the future.

According to Mike Sequeira, Chair of Mathematics at COCC, regression to the mean is more of a description of an event over time than a tool for predicting the future.

"The concept (regression to the mean) really has no bearing on the development of projections," he said. "It's not a forecast model of statistics."

Again, planner Thompson disagrees.

"That's absolutely what 'regression to the mean' is by it's very definition," said Thompson. "Regression to a mean: extreme events that happened in the past will regress toward a historic average in the future. You don't make predictions based on extreme examples. You make them based on an average."

Thompson uses the installation of the sewer system in 2000 as an example of an extreme event that has skewed the rapid growth in recent years since.

Dewey wants the city to stop pushing the comprehensive plan through the process until this question is resolved.

According to Dewey, "The county has acknowledged that they have to do it (the population forecast) over and the urban growth boundary cannot be approved until they solve this issue. The urban growth boundary expansion is built on a house of cards because it's premised on a population forecast that allegedly shows a need to expand. It's totally inappropriate, procedurally, to move ahead."

Thompson fired back, saying, "We don't require a county approval to move forward. Any land use jurisdiction has the right to predict it's population and make planning decisions. It doesn't have to be certified by them."

Brian Rankin, a deputy planner for the City of Sisters, would like more input and cooperation from the critics.

Rankin directed a question to the SFPC: "What methodology for predicting population would you recommend we use?"

Dewey responded, "Use a like situation to make predictions. Take Jacksonville, Oregon, for instance. They installed a sewer system not too long ago.

"Why not look at what's happened to the population in that tourist town?"