*UPDATE* Four Problems with a New Report Blaming Fracking for High School Drop Outs

UPDATE (7/31/15; 10:00am ET): A pair of North Dakota workforce and economic experts have also taken issue with the recent study blaming fracking for increased dropout rates.

In a recent article, Kent Ellis, a workforce development expert who works on career awareness for the North Dakota Petroleum Council and Lignite Energy Council, said that economic opportunity created by any industry could potentially entice young males to drop out of school:

“I don’t think it’s just a boom in energy. A manufacturing boom would have had the same result. Laying it at the feet of the energy industry is really irresponsible. I think any industry would have had the same result. … Some people don’t like the idea of energy development in the state so they’re going to find anything to draw a linkage.”

David Flynn, chairman of the Department of Economics at the University of North Dakota College of Business & Public Administration, agreed the oil and gas industry was unfairly singled out by the study:

“Any economic opportunity that presents in this particular fashion it’s an opportunity for relatively low skill labor.”

Ellis indicated that if there is any evidence of a dropout spike in North Dakota (the state’s graduation rates are among the best in the U.S., by the way) the state’s educational priorities, which lack focus on instilling job skills, are more to blame:

“A good education might be a PhD in philosophy, and then you’ll understand why you don’t have a job. A good education and skills for the workforce are not synonymous.”

Ellis used 2014 data from the U.S. Census Bureau to illustrate his point. In Grand Forks County, home to the University of North Dakota, 92.5 percent of residents have a high school degree, 32.6 percent have a bachelor’s degree or higher, and median household income was $46,745. In Oliver County, home to both coal and oil energy development, 87.5 percent of citizens have a high school degree and just 16.4 percent have a bachelor’s degree or higher, but the median household income $71,250, or 52 percent higher than in Grand Forks County.

Considering nearly 70 percent of U.S. adults do not have bachelor’s degrees, the latter statistic wouldn’t seem to be negative in the big picture should it be applied at the national level. Good-paying jobs are good news for the economy, regardless of education level.

Original post, July 20, 2015

Shale development opponents have blamed fracking for a myriad of calamities over the years, with many of the accusations being as ridiculous as they are unfounded. And with charges of water widespread contamination essentially scratched off the anti-fracking talking points list due to the EPA’s recent landmark groundwater report, the movement has resorted to some rather creative tactics.

Activists recently pushed a new non-peer reviewed, working paper from the National Bureau of Economic Research claiming that kids are dropping out of high school to take high paying jobs in the oil and gas industry.

Bloomberg also produced a headline for a story on the paper – “Fracking Jobs Encouraged American Teens to Become High School Dropouts”— that the Sierra Club would have been proud of. That article quoted the study’s authors, Dartmouth economics professor Elizabeth Cascio and her former student Ayushi Narayan as saying,

“… fracking raises the risk that some workers at the bottom of the skills and education ladder may end up being stuck there, because they made bad schooling choices in a rush to be part of the industry.”

Cascio also claimed that fracking can “leave a potentially permanent negative mark on the skills of its workforce.”

Now, it’s rare that you hear anyone arguing against providing well-paying job opportunities for people regardless of their education level.

But aside from that, the researchers’ claim to have “direct evidence that the effects of fracking on the decision to drop out of high school have been driven by local labor demand shocks that favor the least-educated workers,” is simply not backed up by the facts.  In other words, they actually have no real evidence to claim that shale development has led to a spike in dropout rates.

Let’s have a look at just a few problems that jumped out at us:

Problem #1: Researchers’ use of “commuter zones” inflates dropout percentages with urban data

Rather than using county dropout and oil and gas production data, the researchers use “commuter zones” within major shale plays, which is problematic considering these zones often times extend into urban areas with little or no drilling. More problematic is the fact that urban areas routinely have much higher dropout rates than non-urban areas where a vast majority of shale development occurs.  That can certainly inflate the results to indicate higher dropout rates in drilling areas than actually exist.

For example, one Pennsylvania commuter zone used in the report includes both Washington County, where significant shale development is occurring, and Allegheny County, which features very little shale development and is dominated by the Pittsburgh metro area. Pittsburgh’s graduation rate has historically been very poor, while Washington County’s dropout rate was just 1.1 percent in 2013-14.

Similarly, the prominent Pennsylvania shale gas county of Susquehanna had a dropout rate of just .93 percent in 2013-14, but is lumped in a commuter zone with Binghamton, N.Y.  That’s an interesting move considering that shale development has been banned in New York.  Binghamton High School also has a very poor graduation rate when compared to the rest of New York state, but that can hardly be blamed on fracking.

Further, the county data in several states with shale development show no significant spike in dropout rates since 2006, the year that the report identifies as the start of significant shale development. For instance, just one of Pennsylvania’s top five natural gas producing counties saw an increase in its dropout rate between 2005 and 2013.

Dropout rates in Pennsylvania’s top five natural gas producing counties
                                               2004-05                2012-13
Bradford                              1.8%                      1.4%
Greene                                 2.2                          1.6%
Lycoming                            2.5                          1.9%
Susquehanna                     1.3                          0.9%
Tioga                                   1.2                          1.7%
Statewide                           1.9%                      1.4%

Similar data were found in an analysis conducted by And interestingly, of the 11,620 total dropouts in Pennsylvania in 2012-13, just 555 indicated they quit school because they wanted to work, according to Pennsylvania Department of Education data.

Though dropout information is a bit tougher to find outside of Pennsylvania, there are similar trends in Kern County, California, where a vast majority of the state’s oil and gas development has taken place, and Colorado shale counties such as Weld, Mesa and Rio Blanco, which fall in line with a statewide trend of seeing significant drops in dropouts since shale development began. North Dakota’s graduation rates have also steadily improved since 2009 and continue to be among the highest in the country.

The study’s use of “commuter zones” is a confounding one, considering county-by-county assessments seem to be a more logical and easier-to-understand method of evaluation – which leads us to our next point.

Problem #2: Estimation method of oil and gas production in commuter zones is flawed

Since there is no oil and gas production information available for the specific “commuter zones” used in the paper, the researchers used a complicated procedure to estimate oil and gas reserves in each commuter zone for the purposes of comparing them to each zone’s dropout rate. From the report:

“The first challenge comes in measuring shale oil and gas deposits in a CZ (commuter zone), as there are no existing local (e.g., county-level) estimates. Instead, we must predict a CZ’s reserves based on the location of shale plays and estimates of play-specific reserves, which change annually as new discoveries are made. As a result, our local reserve measures are blunter than they would ideally be and have the potential to be endogenous.”

Such methods seem unnecessary, mind you, considering year-by-year production and/or well count stats are available at the state level, and — in many instances, contrary to the researchers’ claims — the county level and specific shale play level as well. But the researchers focus instead on their estimations, which assume oil reserves are evenly distributed throughout the entire geographical region of each shale play:

“Our preferred approach to predicting CZ reserves relies on the most recent estimates of oil and gas reserves (from 2013) published by the Energy Information Administration (EIA) and the most recent maps of shale plays (from 2011), also published by the EIA. We overlay these maps to counties, separately for oil and gas, and allocate 2013 reserves to counties based on the fraction of each play that they contain, following a process similar to that of Maniloff and Mastromonaco     (2014), who study the local economic impacts of fracking.”

The result of these “estimates” is data that is in no way based in reality. One glaring instance is the two Illinois “commuter zones” (Page 39 of the report) that are included on a map of the zones used for the study. This would indicate that Illinois “reserve” data was used along with dropout data in the report. Illinois, however, has had absolutely no actual shale production thus far. So if the researchers were to have focused solely on Illinois, they would have suggested that fracking has in some way has contributed that state’s dropout rates — despite the fact that fracking hasn’t even occurred in Illinois.

Problem #3: There is no actual evidence that dropouts in the regions studied are taking fracking jobs

The report claims, based on the flawed research methods listed above, that for every 0.1 percent increase in fracking-related employment in the active shale production areas it analyzed, these areas have experienced an increase in the male dropout rate of between 0.3-0.35 percent.

The researchers come to this conclusion based on the notion that because dropout rates among males in non-fracking areas have remained stagnant, while male dropout rates based on their data from “commuter zones” in or near major shale plays have risen, it is clear evidence that fracking is solely to blame:

“Only a quarter of this estimate can possibly be accounted for by in-migration of existing dropouts; the lion’s share of the dropout effect is thus driven by the schooling decisions of local teens.”

However, they provide no actual hard evidence that these dropouts are quitting school to work in the oil and gas industry, acknowledging as much in the report.

“Ideally, we would have high-frequency (e.g., annual) data at the CZ level on dropout rates spanning the introduction of fracking, separately by gender. We would also have similarly frequent and high-quality data on production (by well type) and employment in the oil and gas industry, to confirm that the CZs we classify as having higher fracking potential are indeed the ones where the industry has expanded more rapidly as fracking has spread. The best available data do not meet all of these ideals, but come close.”

Perhaps that’s because a lot of companies require a minimum GED, specifically in Ohio, according to Mark Bruce, Ohio Oil and Gas Education Program Communications Director:

“Based on our experience, on the ground in Ohio, the premise of this study is inaccurate. Through research and conversations with Ohio’s oil and gas industry and employees, the vast majority of oilfield careers in Ohio require at minimum a high school diploma or GED so to draw a correlation between increased oil and gas activity and high school dropout rates, is improper.”

Not surprisingly, Ohio Oil and Gas Energy Education Program Director Charlie Dixon took exception to the study.

Problem #4: Researchers fail to recognize industry commitment to training, education

Interestingly, the researchers’ data showed that wages of male dropouts in shale regions are higher than female counterparts and high school and college educated males. From the study:

“In particular, our identification strategy also uncovers evidence that fracking has increased the relative wages of male dropouts, relative to both female dropouts and males with higher levels of education.”

Again, it’s not often that you hear complaints about jobs that pay too well.  The oil and gas industry certainly does produce high salaries, regardless of education level. Despite the fact that 63 percent of fracking-related jobs are blue collar, as the report points out, the average oil and gas industry salary was $107,200 annually in 2012.

And not only is the industry paying its workers well, even in entry-level positions, it is also making every effort to train current and prospective employees with college programs, technical schools, and high school vocational programs that will allow them to either acquire or expand on the skills needed to work in the industry.

Entry-level training programs such as ShaleNet help train those interested into breaking into the industry in the five high demand upstream occupations, including roustabout, welder’s helper, CDL, floor hand and production technician. Organizations such as the Ohio Oil and Gas Education Program are directly involved in such training and education efforts, as Programs Communications Director Mark Bruce explained in an email to EID:

“OOGEEP works with the industry to provide training for industry workers and to connect them with educational classes and opportunities available. Of the 75 careers available in Ohio’s oil and gas industry, the opportunity for advancement among those positions is great and often attainable.”

As the examples above illustrate, the notion that shale-related positions are temporary, dead-end jobs with no prospect of advancement is simply not true.


The researchers’ primary claim in this report — that fracking is somehow to blame for kids dropping out of school — seems silly on its surface. But closer inspection of the flawed methods and assumptions used to arrive at that conclusion only serves to confirm that they don’t have the facts on their side.


Post A Comment