- Special Projects
Editor’s Note: Four well-known opponents of rail published an opinion piece in the Aug. 21 edition of the Honolulu Star-Advertiser. Civil Beat has identified seven claims worthy of a closer look. This is one of those Fact Checks.
The city’s long-term ridership projections for the rail system are a function of, among other things, the growing population. But rail critics say Honolulu rail planners used an old figure for 30-year population projections even though a new one was available.
In their Aug. 21 op-ed, four well-known opponents made the following statement:
The city has also cherry-picked data. It relies upon a 2004 30-year population forecast even though the 2008 30-year population forecast indicates 100,000 fewer people in 2030 than was previously forecasted.
Asked for the 2008 population estimates the city ignored, one of the piece’s authors, Cliff Slater, provided the following link to Civil Beat:
The city doesn’t dispute that it used an old forecast for a time, but says that it was required to do so. It’s now using the updated forecast, but hasn’t reduced its ridership projections because employment estimates are now higher than they were.
Let’s start at the beginning.
In 2006, when Honolulu was evaluating its transit options, the city produced an Alternatives Analysis. That document used a 2030 population projection of 1,117,300. It said transit ridership under the No Build Alternative would keep pace with projected population growth, and the Fixed Guideway would result in a “substantial increase”1 in transit ridership versus that No Build Alternative.
(Rail opponents raised concerns about the Alternatives Analysis ridership projections that same month.)
The Draft Environmental Impact Statement published in November 2008 continued to use that “old” forecast.
The city didn’t use the new forecast, which came out in 2008 and was last updated in July 2009, until this year. The August 2009 financial plan does not explicitly include a population estimate. The Travel Demand Forecasting Result and Uncertainties Report, published in October 2009, still uses the “old” projections. As does the Final Model Development, Calibration and Validation Report, published that same month.
The Final EIS published in June 2010 continued to use that “old” forecast, too. So there’s at least some truth to the opponents’ claim.
But Honolulu Authority for Rapid Transportation Interim Executive Director Toru Hamayasu says there’s a reason the city didn’t use the new population estimate: It wasn’t allowed to. Hamayasu told Civil Beat Wednesday that federal guidelines require that the city use only population estimates officially adopted by the Oahu Metropolitan Planning Organization, and that OMPO didn’t officially adopt the new model until this year.
The issue of old population projections was raised by Charles Carole, then Vice Chair of the Makiki/Lower Punchbowl/Tantalus Neighborhood Board, in a February 2009 comment on the Draft EIS. City Transportation Services Director Wayne Yoshioka responded to Carole in a June 2010 response that was appended to the Final EIS.
That response goes into detail on the federal guidelines and the implications of the new data from the Hawaii Department of Business, Economic Development and Tourism. We’ll get into the employment figures later in this Fact Check, but you can read the correspondence, provided to Civil Beat by HART Wednesday, here:
In April of this year, a new financial plan produced as part of the city’s request to enter into the final design phase includes the updated population forecasts, citing DBEDT’s 2009 State of Hawaii Data Book as its source.
That table, which matches up with the document Slater referenced, shows the City and County of Honolulu is expected to have 1,017,565 residents by the year 2030.
That final figure is 99,735 less than the city had projected in the 2006 Alternatives Analysis, and backs up the opponents’ claim that the city’s original population forecast is 100,000 higher than a newer forecast.
But the city’s new financial plan for the project uses the new population projections. So it’s not true to say that the city continues to rely on old data. And it’s misleading to say that the city cherry picked its data to support the project. Hence, our grade of half true.
It’s worth noting that the “new” population forecast from July 2009 has already been proven to be inaccurate. It said the city’s population in 2010 would be 911,833, but the 2010 U.S. Census count was 953,207 — a difference of 41,374. If the new actual population were to be extrapolated using the “new” forecast’s annual growth rates, Honolulu would be home to 1,063,737 people in 2030 — almost exactly halfway between the city’s earlier projection and the new projection.
More important, the opponents’ claim about population growth comes in a broader context, under the heading “Unrealistic Cost Estimate and Ridership.” The first part of that section deals with the construction costs and what opponents describe as “an independent study by the highly regarded IMG group.” The opponents say the system will likely cost more than $7 billion to build, not $5.3 billion as the city currently claims.
Civil Beat has already explored that study and the city’s financial modeling in depth for a five-part series published in December 2010. You can read those stories here:
That series, however, did not go into detail on population or ridership projections because IMG did not posit a model to contrast with the city’s projections. In addition to the statement quoted above, the opponents wrote the following in their recent op-ed:
The federal government has compared actual ridership with forecasts in the cities that actually built rail systems and found that these cities overestimated ridership by an average of 41 percent.
When the city prepared environmental impact studies in 1982, 1992 and 2003, it forecast significant increases in bus ridership each time, but ridership declined instead. Yet the city is once again touting wildly optimistic forecasts for rail ridership. These ignore that the most recent population forecast for Honolulu shows that the number of people 20 to 64 years old in the year 2030 is expected to be less than the number today. This age group includes the vast majority of commuters.
But forecasting ridership — a major piece of operating revenue — is far more complicated than just mirroring population growth. In its financial plans, the city listed five other socio-economic variables that, in addition to population, could change the ridership forecasts:
The city did adjust its overall ridership projections between the August 2009 and April 2011 financial plans — increasing the number of projected linked daily trips on both bus and rail combined in 2030 from 281,000 to 283,000.
Hamayasu explained to Civil Beat Wednesday one of the important factors that led the city to leave its ridership estimates unchanged.
“Yes, we didn’t revise the ridership forecast, and here’s the reason: We acknowledge the population went down,” he said. “While the population went down, the employment went up over 5 percent.”
The “old” employment projections for 2030, published in November 2007 according to Hamayasu, were 632,711. That’s the number included in the two October 2009 technical reports2, even though Hamayasu said a higher “new” figure for 2030 employment — 666,191 — came out in September of that year.
“The way that the forecasting model calculates the number of trips is trying to match the population with other activities, such as employment,” Hamayasu said. “So even with less population, if employment goes up, I think it makes sense to think that the amount of travel demand would either stay the same or go up.
“As a result of this sensitivity analysis, we were comfortable sticking with the same number of trips,” he said. “We acknowledge the new population … but we don’t need to change the transit ridership.”
Bottom line: For at least a year after new population projections came out, the city continued to use old figures that opponents say bolstered the case for rail. The city says it had a good reason for that, and its new figures include an increase in employment that leaves the ridership model largely unaffected.
Here are the related Fact Checks:
Between 21 percent and 27 percent, depending on the chosen route.
Both the Travel Demand Forecasting Result and Uncertainties Report and the Final Model Development, Calibration and Validation Report were published in October 2009.