Review of Considine, Watson, and Blumsack’s “The Pennsylvania Marcellus Natural Gas Industry: Status, Economic Impacts and Future Potential” by Matthew Rousu, Zach Zoller, and Dave Ramsaran
This study provides estimates of the economic impact of Marcellus Shale development for 2010 and provides “estimates of the estimates” of the economic impact for 2011 and beyond. The authors use results from surveys that are sent back by gas companies.
Overall, the methods are acceptable. There are some causes of concern, however. Relying on survey responses is a bit worrisome. While it may not be feasible to use other options, it’s quite possible there are reporting errors or that the firms (either intentionally or unintentionally) have misreported information. One has to wonder who within each organization was completing each survey and how much effort they put into it. Or, potentially more worrisome, would the owners of any of the twelve firms that responded realize that this information is going to be used in a PR piece and intentionally overstate the results in order for better press coverage? Economists have long been concerned with issues of “hypothetical bias”, so this is very real concern. With that being said, a study by Hefley et al. (link here) did a detailed analysis of the cost per well, and our back-of-the-envelope calculations find that both that study and this study have similar estimats for the cost per well.
We have some concerns over the multiplier. The authors use Implan, which has been used by many other researchers to estimate economic impacts. The authors find a multiplier of about 2 for the entire state, which might be the right multiplier, but without having the details on how this has been conducted, we have questions. First, does the spending for wells in the center of the state of PA have the same multiplier as the spending for wells located within a 15-minute drive of the state border? If so, this seems concerning. The multiplier for spending should be lower for counties that are on/near the border of other states, as far more spending would “leak” into the neighboring states and wouldn’t stimulate the economy of Pennsylvania. Second, the multiplier higher than multipliers similar researchers used for other studies – and their defense of the higher multiplier seems weak.
Overall, the authors are certainly correct in their statements that the impact of Marcellus shale drilling is enormous – both in terms of direct expenditures and the indirect effects – the question we have is whether the magnitude of the economic impact they’re estimating is correct. Without some answers on the economic impact multiplier, specifically for counties that border other states, we suspect the economic impact multiplier, and hence the overall economic impact, is moderately overstated.
If graded on our "best practices"– how well does this study do?
What the study does well:
2) The study clearly defines the geographic region of the study as it examines the entire state of Pennsylvania.
4) The authors do only count expenditures once. (This seems obvious, but some studies will count expenditures by the firm and then direct spending – which “double-counts” the economic impact.)
7) The study does compare its results, specifically comparing its job multiplier for every million dollars spent by the gas industry to other researchers’ findings. The multiplier used in this paper, it should be noted, is significantly higher than those used in other papers.
8) The study clearly states that it is being sponsored by the Marcellus Shale Coalition and it is published under the name of Penn State University. Also, it states that it received research assistance from the School of Energy Resources and the Center of Energy Economic and Public Policy at the University of Wyoming. If there is bias, one would expect it would be resulting in higher estimates of the economic impact.
What the study does not do well:
1) This study does presents both the dollar impact and the jobs impact. Estimates for 2012 are $14.51 billion and over 180,000 jobs. It would be better if only the jobs impact or the dollar value impact was reported
3) The counter-factual doesn’t seem to be explicitly defined. The authors don’t define whether their looking at total increased economic activity for an area, or economic activity for the people living in the area prior to when the drilling process begins. From reading, it appears that their goal is simply to see all the economic activity from the well, regardless of whom it affects.
5) The study apparently has not been peer reviewed by an outside expert.
6) While the study does use IMPLAN, there are some questions we have regarding the methods – namely how the multiplier differs for regions right near the borders of other states..
· We have concerns both with potential hypothetical bias from survey responses, and with the multiplier that’s being used and whether it’s being varied by counties. Our rough calculations seem to confirm that the authors direct economic activity estimates seem reasonable, but we think the multiplier could be slightly inflated. If that is true, the overall economic impact that’s being reported could be overstated as well.
· We do agree with the authors that is a tremendous economic impact from Marcellus Shale drilling.
· While not really a “fault” of the study, this study does only examine the overall economic impact. It does nothing to discuss who might be benefitting from this development and who might be worse off. A closer look at the differential impact on different groups of people would be useful. (Although, admittedly, it will be more difficult.)
· We don’t comment on the future year’s plans, as there are hundreds of events that could change plans for 2015, not to mention 2020. Our review is examining the estimates the authors provided for 2010.