“My hope would be that any reporter who is looking at the facts would take the time to confirm that the most realistic estimates are this might create maybe 2,000 jobs during the construction of the pipeline — which might take a year or two — and then after that we’re talking about somewhere between 50 and 100 [chuckles] jobs in a economy of 150 million working people.”
— President Obama, interview with The New York Times, July 24, 2013
This column has been updated, with the rating reaffirmed.
We are always interested when the president directs reporters to look at the facts.
Readers may recall that in 2011 The Fact Checker had looked deeply at the question of the number of jobs that might result from building the Keystone XL pipeline. We labeled it “a bipartisan fumble,” knocking lawmakers on both sides of the aisle for uttering greatly inflated job estimates that in one case even topped 100,000.
But now here’s the president, tossing out a rather low figure (“maybe 2,000” during the construction phase) and then chuckling that it would only be “50 to 100 jobs” after that.
When we had looked at this before, we concluded that all such estimates are subject to guesswork, but a mainstream estimate appeared to come from the State Department — 5,000 to 6,000 construction jobs in a year. TransCanada, which would build the pipeline, had its own, somewhat similar estimate for the two-year project — 13,000 jobs, or 6,500 per year.
The numbers get fuzzier after that, because thousands of “spin-off” jobs (suppliers, and suppliers of suppliers) get added into the mix. Believe it or not, such claims can get far afield, adding in dancers, dentists, clergy, bartenders and the like who supposedly receive jobs because of a big construction project. But at the same time, there clearly is also a second-order effect of some sort.
So how does the president end up with such a low figure?
The White House would not explain the president’s math, except to point to an anodyne statement made by spokesman Josh Earnest at Monday’s daily news briefing, after he was asked about the president’s jobs estimate, which was published in the Sunday edition:
“There are a range of estimates out there about the economic impact of the pipeline, about how this pipeline would have an impact on our energy security. There are also estimates about how this pipeline may or may not contribute to some environmental factors. So there are a range of analyses and studies that have been generated by both sides of this debate.”
Our colleague Juliet Eilperin suspects that the president is relying on an estimate generated by the Cornell University Global Labor Institute, which opposes the pipeline project. Cornell figures each segment of the pipeline requires 500 workers per segment. The southern leg of the pipeline is now nearly complete, so that means 10 segments are left. That translates into 5,000 jobs over two years, or 2,500 a year.
Meanwhile, because part of the pipeline is complete, the State Department has revised downward its estimate of the construction jobs to 3,900 jobs per year over a one-to-two-year period. That’s still a higher figure than the one generated by opponents.
The State Department also says the project could “potentially support approximately 42,100 average annual jobs across the United States over a one-to-two-year period.” State said the employment would translate into about $2 billion in workers’ earnings, $3.3 billion in construction and materials costs and $67 million in state and local taxes. That sounds like real money and quite a few jobs, at least in the short term.
Still, echoing what the president said about operation of the pipeline, State says that “operation of the proposed project would generate 35 permanent and 15 temporary jobs, primarily for routine inspections, maintenance, and repairs. Based on this estimate, routine operation of the proposed pipeline would have negligible socioeconomic impacts.”
Ordinarily, we would expect the president to cite an estimate from his own State Department, rather than a think tank opposed to the project. (Note to President Obama: When researching such matters, reporters generally look askance at estimates produced by advocates or foes of a particular issue.)
Of course, perhaps the president just took State Department estimate of the construction jobs and divided it in half, to come up with an (incorrect) yearly figure.
But that doesn’t make much sense either, because the White House routinely claimed the job gains created by the stimulus by adding up the number of “person-years” — in other words, one person employed per year. That’s how the White House could claim 3 million jobs were saved or created by the stimulus through 2012. (See Table 12 of this White House report.)
Thus, using the White House’s stimulus math, the president should be saying Keystone XL would create as many as 7,800 construction jobs. (Note: please see update below.)
The Pinocchio Test
Predictions of possible jobs are always fraught with complications, guesstimates and fuzzy math, so they often should be taken with a grain of salt. No one really knows exactly how many jobs will be created. So maybe the president is right to be skeptical.
But the president shouldn’t pick and choose how he cites job-creation numbers. Perhaps he is tipping his hand on what he secretly thinks of the Keystone XL by citing a low-ball figure, generated by the pipeline’s opponents, but he should stick to using the official government estimate. (His 2,000-job figure is actually slightly lower than the Cornell estimate.)
Otherwise, the president ironically seems to be signaling that even his own government does not produce the “most realistic” estimate that should be used by reporters.
UPDATE: The National Resources Defense Council, which opposes Keystone XL, has posted a blog post challenging this analysis and arguing that the president’s math is correct. We find it curious--and suspicious--that the White House has refused to explain the president’s reasoning, thus leaving it to outsiders to parse his words.
As always, the burden for proof remains with the politician. If the White House decides to explain the president’s remarks, and his math adds up. we will revisit this ruling.
UPDATE, August 2: After carefully reviewing the State Department document again, and discussing it with officials, we believe there is a fundamental confusion about what the document says concerning the jobs creating by Keystone.
We have may added to the confusion with our paragraph above about the stimulus accounting, since the calculations done by the State Department are not done under the same method as the stimulus accounting. So this was a facile observation, mixing apples and oranges.
But that does not make the president’s statement correct. In fact, it becomes even more mystifying.
Table 4.10-3 in the State Department draft environmental statement lays out the number of construction workers per location, with a construction period for each state. The report has a formula for determining the number of workers in a year: Multiply the number of workers times the construction period and divide by 52 (weeks). Under this formula (which appears in footnote 4 on page 6), it shows:
Montana: 4,000 construction workers for an average of 19 weeks = 1,462 workers
South Dakota: 3,500 construction workers for an average of 20 weeks = 1,346 workers
Nebraska: 2,700 construction workers for an average of 19.5 weeks = 1,013 workers
Kansas: 200 construction workers for an average of 33.5 weeks = 129 workers
That adds up to 3,950 workers.
People appear to be getting caught up in the question of whether the project is over two years or not. But our reading of the report –and this table—is that it would be incorrect to adjust these numbers for the length of the project (which rather than two years is more like two construction seasons.) Thus, there are two ways to express what the State Department says:
10,400 construction workers will get jobs that generally last for just under half a year.
Or, on an annual basis, 3,900 workers will get jobs.
(In case you are wondering why the detailed breakdown yields 3,950 workers and State officially says 3,900, it appears State simply did a quick and dirty calculation of 10,400 x 19.5 weeks divided by 52. That’s not quite accurate because the Kansas leg of the project takes an average of 33.5 weeks, which is much longer than 19.5 weeks. Kansas would not employ a lot of workers but the result of State’s calculation is that the number of workers in the overall annual calculation is reduced by 50.)
Put another way, the “yearly” figure is not a per-year figure; it just squeezes all of the work into 52 weeks and calculates how many workers would be employed for an entire year.
One cannot simply declare that the project will last two years, and thus that means about 2,000 jobs a year. Similarly, you could not say the project lasts six months (which is closer to the truth, as it is a series of six- month jobs) and thus it employs 7,800 workers.
The White House surely understands this, which is why officials presumably have remained so unwilling to explain the president’s reasoning. We reaffirm our original ruling.
Check out our candidate Pinocchio Tracker