HOUSTON -- Two economic modeling firms believe the proposed final phases of the Keystone XL pipeline project will generate thousands of jobs, but not tens of thousands of direct jobs, as many of its proponents assert.

And despite opponents' objections, Keystone XL will probably not affect gasoline prices in the Midwest or anywhere else in the United States, though it could help lower the cost of crude oil globally.

In a study made public this week, Regional Economic Models Inc. (REMI) of Massachusetts and Energy & Water Economics (E&WE) of Tennessee attempt to bridge the variances between two competing Keystone XL studies, one by consultants to the company building it and the other by Cornell University's Global Labor Institute (GLI). The firms find fault with both of those assessments.

The Perryman Group (TPG), a Texas financial analysis company hired by TransCanada to study the pipeline's economic impact, predicted earlier that Keystone XL could help generate up to 59,500 direct jobs for the two years it would take to build it. Cornell researchers have arrived at a much lower number, and even said the impact on jobs could be negative if Keystone XL causes gasoline prices to rise, particularly in the Midwest.

REMI and its partner say both are wrong. Running the data used by GLI and TPG, they see up to 16,000 direct jobs created nationally for two years, most in states where Keystone XL would be built. About 800 permanent jobs would be supported in later years, the companies predict.

The firms also believe that the pipeline would have little impact on gasoline and other refined product prices in the Midwest, suggesting a net economic benefit to the United States, though one more modest than what TPG predicted. Studying the primary literature on the topic, the researchers could find no evidence that Keystone XL would raise gasoline or diesel fuel prices by itself.

"We can consider no sure energy market impact based on refined product price or crude supply disposition uncertainty," writes Scott Nystrom at REMI and William Wade at E&WE, the two authors of the report.

If anything, crude oil prices would probably decrease after Keystone XL came fully online, the authors say, though they would not predict by how much. A lower oil price may or may not cause refined fuels prices to drop. This is because Canadian oil sands crude would enter into competition with Middle East oil suppliers for the business of refineries in the Gulf of Mexico region.

Opponents of Keystone XL say the project would take crude supplies away from Midwest refineries, causing energy prices to rise, but the authors say that rising total crude projected from the oil sands region and the Bakken Shale will leave refineries with ample supplies.

"The effect of increased competition between Canadian and Arab suppliers suggests that crude input prices may decline during this decade but without effect on consumer product prices," they say. "A central result of the KXL is that it will allow Canadian crude one more way to reach growing world refinery demand for crude inputs."

The researchers say the dynamics influencing refined product pricing in the United States are too numerous and complex to be much affected by Keystone XL's completion alone. Experts attribute the current higher gasoline prices in the United States to a strong crude oil price, driven by fears of disrupted Iranian exports, and the anticipation of lost refining capacity in the Northeast.

Disagreement on numbers

The REMI, E&WE and TPG analyses all agree that building Keystone XL would generate both direct and indirect economic boosts to the United States, but two disagree on the total numbers by a wide margin.

TPG reported that, nationally, up to 250,000 total jobs could be created by the Keystone XL pipeline construction and the other economic benefits it would bring by 2015. REMI and E&WE put that figure at closer to 60,000 jobs in total, in both direct and indirect employment, by that same year.

TPG saw the project adding roughly $29 billion to the U.S. gross domestic product (GDP). REMI and E&WE put the number at about one-fifth of that, or around $5.3 billion in added value to the GDP.

A REMI representative said the new study was done by using the same inputs employed by TPG, but then running them through a different economic model using a different methodology. They also removed from the equation what they described as false assumptions initially taken by GLI that cast the Keystone XL in too negative a light, leading to the middle-road estimates the two economic modeling firms have since published.

"The Perryman Group estimated what we thought were high numbers, and GLI did not attempt to re-estimate them (for lack of the proper model and/or expertise)," Nystrom explained in an email. "We tried to 'fill the gap' by re-estimating numbers in our model with a combination of Perryman and GLI's inputs."

REMI says its system is widely employed by governments, utility companies and academic institutions.

TransCanada recently made a formal announcement that it would move forward with building a segment of Keystone XL that would connect the Houston and Port Arthur refining zones with the massive crude oil storage complex at Cushing, Okla. The entire project is on hold pending a final decision by the Obama administration on whether to allow the pipeline to be built across the U.S.-Canadian border.