Dueling Studies – Conflicting Job Numbers for the Keystone XL Pipeline

David A. Foti

Professor of Economics,

Lone Star College (Houston, Texas)



Dueling Studies – Conflicting Job Numbers for the Keystone XL Pipeline

David A. Foti, FRM, PMP

Professor of Economics, Lone Star College (Houston, Texas)

President and Founder, Foti LLC



The debate over whether or not to construct the Keystone XL Pipeline goes deeper than just jobs.  Opponents speak to their environmental ideology – concerned with the potential pollution from leaks and an acceleration of global warming.  Proponents argue in terms of free markets and energy security.

So while the root of the debate is about conflicting visions for America’s energy future, the ground tactics required by the political reality of stubborn elevated unemployment has cast the Keystone Pipeline decision in terms of job creation.  Interestingly enough, both sides are as far apart on their job creation estimates – which should be fact based – as they are in their ideology.

This paper looks at the two main studies being used in the debate and evaluates each on its relative merits.


Keystone XL Pipeline

TransCanada proposed the original Keystone Pipeline in 2005 and completed it in 2010 and 2011 (Phases 1 and 2, Figure 1).  Given rising oil prices and oil sands production, TransCanada proposed an expansion to the original Keystone Pipeline, called Keystone XL, in 2008[1] (Phases 3 and 4, Figure 1).

The proposed Keystone XL Project consists of a 1,700-mile crude oil pipeline and related facilities that would primarily be used to transport Western Canadian Sedimentary Basin crude oil from an oil supply hub in Alberta, Canada to delivery points in Oklahoma and Texas. The proposed Project would also be capable of transporting U.S. crude oil to those delivery points. The proposed project could transport up to 830,000 barrels per day[2].

Figure 1


Source:  http://www.transcanada.com/keystone_pipeline_map.html


After initially supporting the project in its first and second draft environmental reports[3], on January 18, 2012 the U.S. Department of State under intense political pressure from the environmental lobby and ongoing concerns about the proposed route through Nebraska reversed itself and recommended that “the presidential permit for the proposed Keystone XL Pipeline be denied and, that at this time, the TransCanada Keystone XL Pipeline be determined not to serve the national interest”[4].  The decision to revisit approval was set to take place in the first quarter 2013, after the 2012 presidential elections.

Canadian Prime Minister Stephen Harper said he was “profoundly disappointed” and spoke of the need to diversify Canada’s oil industry.  Ninety-seven percent of Canadian oil exports currently go to the U.S[5].  The Canadian government reacted to the Obama administration’s decision to delay a decision on the pipeline by pushing forward an alternate pipeline, the Northern Gateway Pipeline.  The Northern Gateway Pipeline would move crude west for export to serve growing Chinese demand (Figure 2).

Figure 2

Proposed Route for the Northern Gateway Pipeline


Source:  http://www.northerngateway.ca/assets/pdf/General%20Project%20-%20Regulatory/NGP-FS-01-001_Northern%20Gateway%20Project%20Overview.pdf


“I think what’s happened around the Keystone is a wake-up call, the degree to which we are dependent or possibly held hostage to decisions in the United States, and especially decisions that may be made for very bad political reasons,” Harper told local press[6].

On February 27, 2012 TransCanada announced that it would proceed with the Cushing to Houston portion of the project.  This section of the proposed pipeline does not cross an international border into the United States, and so is not part of the Presidential Permit process.  Separately, TransCanada gave the State Department advance notice of its intention to submit a new application for the cross-border segment of the Keystone XL pipeline, from Canada to Steele City, Nebraska, once a new route through Nebraska has been identified[7].

The Perryman Group Study

TransCanada commissioned The Perryman Group (TPG) to evaluate the impact of developing the Keystone XL on U.S. business activity.  The Perryman Group is a small economic consultancy located in Waco, Texas.  According to its website, TPG has a staff of seven, including its bookkeeper and two economists.  Dr. Ray Perryman is its founder and has served as a professor of economics at Baylor University and Southern Methodist University.  TPG issued its report in June 2010.  The report was tremendously supportive of the Keystone XL project and estimated that it would result in 118,935 person years of employment from construction and development and with the multiplier effect create an additional 250,000 to 500,000 new jobs in the U.S[8].

TransCanada posted the report on its website and supporters of the project began to use the TPG reports conclusions as public talking points.  For example, Senator Kay Bailey Hutchison (Republican, TX) said to the Senate that the project "promises 20,000 immediate jobs and 118,000 spin-off jobs."[9] Exxon has created and is running a TV commercial using the 500,000 job figure[10].

Additionally, some prominent commenters have used TPG’s claim that Keystone XL would create 200,000 jobs[11].

Cornell University Study

The Cornell Study has been the primary source used to question the results of the TPG study by analysts and the media.[12]

In September 2011, Cornell University’s Global Labor Institute (Cornell) issued its own study that was highly critical of the TPG study.  The Cornell study challenged TPG’s conclusions on a point-by-point basis.

According to its website, “the Cornell Global Labor Institute was established in 2005 to work with trade unions and one of the main goals of the Institute is to help union officers, staff and activists gain a deeper understanding of the policies and institutions that shape today’s world.”[13]

The co-authors of the study, Sean Sweeny and Lara Skinner, are not economists and do not have a background in the energy industry[14].  Their expertise lies in union and labor studies.

The Cornell study asserted that the TPG study was deeply flawed and “lacked adequate data and detail to render the report so opaque as to make meaningful review impossible”[15].  The Cornell study concluded, based on their own analysis, that the Keystone XL project would create very few American jobs and that the benefit of these jobs would probably be more than offset by the environmental damage attributable to the pipeline – viz. leaks and an increase in global warming.  The specific differences between the two studies are discussed in detail in the following section.


Comparing the Studies

Job creation from Keystone XL is most likely to result from the following:

  1. Direct expenditures related to planning and construction
  2. Multiplier effect of the direct expenditures
  3. Effects of lower and less volatile oil prices
  4. Permanent jobs created to operate and maintain the pipeline
  5. Property taxes paid on a continuing basis
  6. Increase in Canadian imports of U.S. goods and services

Let us take each of these categories in turn, and examine what TPG and Cornell studies say about each, as well as other sources.

Direct expenditures related to planning and construction

TransCanada states that 13,000 direct construction jobs and 7,000 associated manufacturing jobs will be created by the construction[16].   The 13,000 estimate for direct construction jobs fails to note that the majority of these workers would only be employed for 6-9 months each[17].  The U.S. State Department more properly estimates direct construction jobs between 5,000-6,000 for a period of 1-2 years[18].

TransCanada does not provide any information to support its estimate of 7,000 associated manufacturing jobs.  The largest cost input for the pipeline is the steel pipeline – i.e. the ‘pipe’ in pipeline.  According to the Cornell study, TransCanada “imported almost all of the steel pipe need for the U.S. portion of Keystone Phase 1” from India.  Additionally, Cornell asserts that TransCanada has already signed contracts for the steel pipe for Keystone XL to be produced in Canada and India[19].

While the steel pipe may be ‘finished’ in the U.S., the substantial degree to which imported steel pipe will be used argues against a 7,000 U.S. job growth spurt from associated manufacturing.

Multiplier effect of the direct expenditures

The TPG report does not break down its construction job estimate into direct and multiplier effects.  Instead it states a combined value of 118,935 total job years created[20].  Significantly, TPG provides no detail as to how this figure is calculated, but relies instead upon a “black box” model that it says is proprietary.

Even with the limited data TPG provides on its estimate, the Cornell report challenges the few details provided in their report.   TPG seems to include in its analysis both the impact of the Cushing line which has already been built for $1.2 B, and $7 B for new construction to be spent on Keystone XL.   Of this $7 B,

  • $1.6 B will be spent on Canadian portion (i.e. no U.S. jobs)
  • $1.4 B has already been spent, mostly on design, permitting, and material inputs

After stripping these extraneous pieces out of TPG’s total estimated spend of $8.2 B ($7 B for Keystone XL + $1.2 B for the already completed Cushing line), the new total is $4 B for U.S. spending[21].

Even assuming TPG’s black-box multiplier model is correct, it is using a core input (i.e. incremental spend) that is highly inflated.  Given that, it is reasonable to look upon TPG’s estimate of 118,935 total job years created with a substantial degree of skepticism.

Using a corrected spend number and comparable multiplier effects from other pipeline projects, Cornell estimates the multiplier effect to be in the range of 22,000 jobs for two years[22].

Effects of lower and less volatile oil prices

TPG really gets its multiplier effect into high gear for this section.  TPG’s estimate for job creation as a result of lower and less volatile oil prices from increased Canadian supply via Keystone XL ranges from 250,000 – 550,000 new jobs.

TPG does not provide any details as to how much they think oil prices and volatility will decrease, so it is problematic to evaluate their results.   However, from previous TPG reports – in this case for a proposed wind farm – we can get some insight into what types of jobs they expect to be created,

Among the list of jobs that would be created: 51 dancers and choreographers, 138 dentists, 176 dental hygienists, 100 librarians, 510 bread bakers, 448 clergy, 154 stenographers, 865 hairdressers, 136 manicurists, 110 shampooers, 65 farmers, and  1,714 bartenders.[23]

Moreover, the Cornell report actually seems to believe that increased oil supply will kill jobs.  They point to a study[24] that shows that the Keystone XL will decrease basis differentials between the Midwest and the Gulf Coast resulting in higher gasoline prices for Midwesterners thus killing jobs.  Of course, by the same logic the lower prices experienced by Gulf Coast residents will create jobs.

Basic macro-economic theory tells us that natural resources are a key factor of production, and that cheaper natural resources shift the aggregate supply curve to the right, thus increasing GDP and lowering inflation.  So while neither the TPG nor Cornell study provide much insight on the impact of lower and less volatile oil prices, it is safe to assume that the effect will be positive and non-trivial.


Permanent Jobs and Property Tax Revenue

TransCanada has not provided specific information on how many permanent jobs will be created to operate and maintain Keystone XL, but most estimates put it in the range of 50-100.  Even a Keystone VP said permanent U.S. jobs would be about the number of employees at one or two Wal-Mart stores[25].

TransCanada reports that it expects to pay $5.2 billion in property taxes[26] over the 100 year life[27] of the pipeline, or $52M per year.  Assuming half of this amount is spent on personnel and a $150K loaded civil servant salary that would add approximately another 170 jobs initially, declining at the rate of inflation thereafter.  However, there is no guarantee that states will spend this incremental revenue on workers.  Furthermore, some of these property taxes have an abatement period up to ten years so 170 direct jobs is overly optimistic. 

Increase in Canadian imports of U.S. goods and services

While neither study examined potential for increased U.S. exports as a result of increased Canadian disposable income, the American Petroleum Institute (API) did opine.  Rayola Dougher, senior economic advisor at the American Petroleum Institute, said that “for every dollar the U.S. would spend buying Canadian oil, Canada would spend 90 cents buying U.S. goods and services.”

She further states that, “Crude is going for about $100 a barrel or so. So just imagine 700,000 barrels a day, initially, at $100 a barrel, and then 90 percent of that money coming back to the United States," she poses. "You very rapidly can get to about 500,000 jobs within the next 20 years or so."[28]

This analysis is easily challenged. First off, Canada’s marginal propensity to import from the U.S. is 16% not 90% as asserted by Dougher.

  • Canada GDP = $1,577 billion[29]
  • U.S. exports to Canada = $248 billion[30]
  • Canadian marginal propensity to import from U.S. = 16%

Secondly, Dougher assumes that this oil would not be otherwise sold if not for the Keystone XL project.  The Canadian government’s aggressive push for additional markets with projects such as the Northern Gateway Pipeline proves otherwise.


Unfortunately, the studies produced to-date regarding job created from the Keystone XL project confuse as much as enlighten the public debate.  It seems as if the TPG study was intended to produce a certain result, and to that end, engages in opaque analytics and black-box predictions.  The Cornell study, while more professionally produced, is also tainted with the preeminence of its environmental movement ideology.

While neither study is the robust independent analysis that policy makers would ideally use to make an informed decision, each does provide usefully data points to advance the conversation.  After cutting through all the noise of the two studies, it is clear that the Keystone XL will directly create U.S. jobs in the order of 5,000 to 6,000 over the short-term, and 100 to 200 over the longer-term.  Furthermore, the 500,000 barrel per day increase in import capacity[31] from Keystone XL will increase America’s energy security and should have a positive impact on reducing U.S. oil prices and volatility which would be supportive of  higher U.S. GDP and employment over the long term.



[1] Hovey, Art (2008-06-12). "TransCanada Proposes Second Oil Pipeline". Lincoln Journal-Star (Downstream Today). http://www.downstreamtoday.com/news/article.aspx?a_id=11336.

[2] http://keystonepipeline-xl.state.gov/documents/organization/181184.pdf

[3] http://insideclimatenews.org/news/20111019/epa-lisa-jackson-state-department-controversial-keystone-xl-environmental-review-hillary-clinton-transcanada

[4] http://www.state.gov/r/pa/prs/ps/2012/01/181473.htm

[5] http://fuelfix.com/blog/2012/01/29/thwarted-on-us-oil-pipeline-canada-looks-to-china/

[6] Ibid.

[7] http://www.ens-newswire.com/ens/feb2012/2012-02-27-01.html

[8] The Perryman Group, “The Impact of Developing the Keystone XL Pipeline Project on Business Activity in the US”, June 2010, pages 4 and 6.

[9] http://www.npr.org/blogs/itsallpolitics/2011/12/14/143719155/just-how-many-jobs-would-the-keystone-pipeline-create

[10] http://www.youtube.com/watch?feature=player_embedded&v=PObSVK8qyXc

[11] http://mediamatters.org/mmtv/201112200013

[12] http://www.cbsnews.com/8301-505123_162-57361212/keystone-pipeline-how-many-jobs-really-at-stake/


[13] http://www.ilr.cornell.edu/globallaborinstitute/about/

[14] http://www.ilr.cornell.edu/globallaborinstitute/contact/

[15] Cornell University Global Labor Institute, “Pipe Dreams? Jobs Gained, Jobs Lost by the Construction of the Keystone XL”, September 2011, page 18

[16] Keystone Connection Special Edition February 2012

[17] US State Department’s Final Environmental Impact Statement (FEIS), Socioeconomics, Section 3.10-54

[18] FEIS, Socioeconomics, 3.10 - 55

[19] Cornell University Global Labor Institute, “Pipe Dreams? Jobs Gained, Jobs Lost by the Construction of the Keystone XL”, September 2011, page 12

[20] The Perryman Group, “The Impact of Developing the Keystone XL Pipeline Project on Business Activity in the US”, June 2010, page 22.

[21] Cornell University Global Labor Institute, “Pipe Dreams? Jobs Gained, Jobs Lost by the Construction of the Keystone XL”, September 2011, page 6

[22] Cornell, page 24.

[23] http://www.washingtonpost.com/blogs/fact-checker/post/keystone-pipeline-jobs-claims-a-bipartisan-fumble/2011/12/13/gIQAwxFisO_blog.html

[24] Verleger, Phil. “If gas prices go up further, blame Canada,” Star Tribune. May 13, 2011.

[25] http://missoulian.com/news/opinion/mailbag/republicans-cry-crocodile-tears-for-keystone-xl-jobs-bust/article_25fa9b6c-55d1-11e1-9551-001871e3ce6c.html

[26] http://www.transcanada.com/docs/Key_Projects/know_the_facts_kxl.pdf

[27] The Perryman Group, “The Impact of Developing the Keystone XL Pipeline Project on Business Activity in the US”, June 2010, page 11.

[28] http://onenewsnow.com/Business/Default.aspx?id=1513148

[29] http://data.worldbank.org/country/canada

[30] http://www.ustr.gov/countries-regions/americas/canada

[31] http://www.downstreamtoday.com/news/article.aspx?a_id=11890&AspxAutoDetectCookieSupport=1



Click to view a printable version of this article.