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.
Proponents have argued that a major boon from the proposed rail project would be the creation of local jobs. Critics say the employment benefits have been overstated.
In their Aug. 21 op-ed, four well-known opponents started doing the math. Under the heading “Exaggerated Job Creation,” the four make the following statement:
The city initially claimed that rail would create 17,000 new jobs during the construction phase, but later lowered its estimate to 10,166, without explanation. Even this number is pure fiction.
The $483 million construction contract went to Kiewit, which said it needed 350 workers to build the first segment. The same workers would probably end up building the remaining segments, because the plan is to build the system in segments, not all at once.
An Italian company, Ansaldo, expects to receive more than $1 billion for providing and maintaining the trains and rail system. It is promising “300 local jobs for local people.”
If you are counting, we have identified 650 new jobs. The city has yet to identify the other 9,516 that it has promised.
Asked to explain if the opponents believe the rail project will only create 650 local jobs, or if those were the only two sources that have identified how many workers they plan to hire, one of the piece’s authors, Cliff Slater, provided the following statement to Civil Beat:
Those are the only two sources. However, the public relations incentives are for the contractors to increase the projected jobs, not decrease, from what they believe is true.
In trying to pin down the credibility of the city’s projections, we first looked to how rail planners came to the numbers in the first place. The city estimates came from a widely used economic model from the U.S. Department of Commerce’s Bureau of Economic Analysis.
The Regional Input-Output Modeling System — more commonly referred to as RIMS II — is used to assess the economic impacts of a broad range of public and private projects, policies and events. Officials use RIMS II to speculate about how everything from natural disasters to the construction of major infrastructure will affect the economy.
Economists using the RIMS II model take national labor data from various industries and regionalize it by determining what kind of labor is available in a given area.
As for the numbers themselves, here’s how rail spokeswoman Jeanne Mariani-Belding explained the discrepancy between the 17,000 and 10,000 estimates cited in the commentary by rail opponents:
“More jobs are generated during construction years than during the planning phases and the overall yearly estimate is roughly 10,000 each year – with peak construction years hitting more than 17,000. The city has always said that construction of the rail project would generate an estimated 10,000 direct and indirect jobs that are both in and outside the construction industry. The 10,000 estimate comes from taking the job estimates numbers in the years of construction and averaging them.”
City officials definitely have used the 17,000 number in the context of selling the project. For example, in February 2010, former Mayor Mufi Hannemann said the project would create 17,000 jobs in 2013, but he also at the same time gave much lower estimates for other years. Civil Beat could not find, nor could rail opponents point us to an example where the city lowered its projection from 17,000 to 10,000.
Still, an average of 10,000 new jobs per year? That’s a work force nearly the size of the entire City and County of Honolulu, and a number that ought to make anyone skeptical.
One valid question that Slater raised in an email to Civil Beat about this Fact Check is that the city doesn’t take into account the way the local tax approved to pay for rail negatively affects the local economy, and job creation.
“Doesn’t it trouble you that the city has acted as though taking billions of dollars out of the private sector (in taxes to pay for rail) would not cost any jobs?” he wrote.
“Common sense alone tells you that removing billions from the private sector will have a significant adverse impact on the local economy,” he wrote. “Why do you suppose the city has not mentioned that aspect of the jobs question, or factored it into a projection on net job creation?”
A closer look at economic models like RIMS II provides at least some clarity.
First things first, the 10,000 figure refers to full-time job equivalents. In other words, we’re not talking about adding 10,000 jobs every year until we have 50,000 at the end of five years. Instead, if the prediction is accurate, it would be an average of 10,000 jobs each year.
Something else that falls by the wayside in the city’s “10,000 new jobs” sound bytes is the fact that a significant portion of the jobs will not be local.
In an input-output model, economic forecasters look at all kinds job creation even peripherally tied to rail. For example, they count the jobs that they believe the rail project will create at a local gravel company but also the jobs that the rail project could create at a foreign rail-car manufacturer.
Some of the jobs that forecasters count are not at all related to rail. Forecasters assume that — given the job creation that rail spurs — those now-employed people will go out and spend more money. As economic demand goes up, the argument goes, so too does hiring.
Here’s how the city explained it in the draft Environmental Impact Statement, which it did not revise for the final version:
“Based on construction cost estimate and state-specific employment multipliers, construction-related employment was estimated for direct, indirect and induced employment.”
This table, again from the draft EIS, explains the multipliers used to determine how many jobs would be created per $1 million spent:
The multipliers are produced by the Hawaii Department of Business, Economic Development and Tourism, and are industry-specific. “Direct” jobs are jobs created in the construction industry; the 650 jobs identified by Kiewit and Ansaldo would fall into this camp. Indirect jobs are jobs created in other fields (for example, restaurants that serve construction workers); induced jobs are created when the state’s economy grows as a result of the construction.
Because of forecasters’ reliance on indirect job creation, economists do not identify which specific jobs they expect will be created. A 2009 fact sheet from the city estimated that 48 percent of the jobs created by rail would be in the form of a combination of indirect construction and induced jobs from increased consumer spending.
The particularly frustrating thing about economic modeling is that it’s difficult if not impossible to verify accuracy of predictions even after the fact. (Just look at the wrangling in Washington over the extent to which the American Recovery and Reinvestment Act of 2009 created jobs.)
It’s important to note, as Civil Beat previously reported, that University of Hawaii economist Sumner La Croix last year said the city should reduce its job estimates by about one third, based on new data showing lower consumer spending due to unemployment and wage cuts.
“These multipliers the city uses are way too high,” La Croix said. “We have newer numbers.”
Slater told Civil Beat “the city’s numbers are ‘fictional’ as opposed to ‘factual,’ because they are the product of speculative theory—that is, to produce these numbers the city apparently did little more than plug expenditure figures into a forecasting model, which, as you acknowledged in your email, requires speculation.”
It’s fair that rail opponents are skeptical about the 10,000-job estimate, especially given how difficult it is to count tangible local job opportunities directly tied to rail.
The city has used both 17,000 and 10,000 as figures for job creation, but to make it seem that it lowered the total from 17,000 to 10,000 is misleading. The opponents provide no evidence1 that the city lowered its estimate, nor could Civil Beat find any. The 17,000 figure has been used as a peak figure, while the 10,000 figure has been used as an average.
While we cannot determine whether the Commerce Department’s models are accurate, the city did not pull its job-creation estimates out of thin air.
It used a widely accepted model to estimate job creation.
Unfortunately, the claim about the total number of jobs that will be created is one that may never be verifiable, even during construction. Jobs created by Honolulu’s rail project can occur anywhere in the world, and some are impossible to tie to the project because they are expected to be generated as part of the greater economic impact of the project, something that is rarely made explicitly clear by city officials.
Opponents are correct that the city has identified a limited number of new jobs so far. But it’s not fair to give the impression that the city should be able to identify all the jobs the project will create. That’s not the way these job models work. Indirect and induced jobs, which make up 60 percent of the city’s claim, can never be identified.
It is accurate, however, to say that the city has left the impression that the 10,000 jobs that it says will be created will be in Hawaii, which is not the case. The city hasn’t always taken the extra step to make clear that its 10,000 number is an overall global total.
Bottom line: The opponents have some basis for saying the job creation estimates are exaggerated. They are correct that specific jobs haven’t been identified, but that’s an unrealistic expectation. The rest of their claims are not supported by facts.
That is why our grade in this case is half-true.
Read opponents’ response to this Fact Check: Click here.
Here are the related Fact Checks:
In an email response to a query from Civil Beat about the basis for the claim that the total jobs created initially was 17,000 and then was lowered to 10,000, Slater sent the following link to a Star-Advertiser article. However the article does not quote any government official saying that the number had been 17,000, nor was that mentioned in the press release or in Civil Beat’s three articles on the groundbreaking where the alleged statement occurred.