I appreciate the opportunity to briefly respond to the paper issued last week by researchers from Ohio State’s Dept. of Agricultural, Environmental, and Developmental Economics.

In that report, authors Partridge and Weinstein take issue with the methodology employed in a study we prepared in September for the Ohio Oil and Gas Energy Education Program (OOGEEP). Throughout the text, the authors both state and imply that our analysis isn’t “independent,” presumably because our work drew on data and funding provided by Ohio energy producers. It also suggests that the methods we use to arrive at our conclusions are “old” and “unreliable.” Neither assertion is accurate.

It’s important to note that the primary intent of our research project is to prepare Ohioans for what is expected to be a significant upsurge in activities associated with energy development in the state. This study was not prepared for the purpose of seeking additional research grants, nor does it advocate in favor of one particular energy source, or one particular energy policy, over another.  The purpose of the study is to inform users about the general costs and benefits of potential Utica development.  It is in this sense that it can be used as an aid in decision-making, providing information on the extent of support a community or region might be able to provide when planning for different development possibilities.

It’s also worth pointing out that we only modeled what the first five years of Utica development might look like, drawing on the best information available to us. Over time, it is of course true that direct returns from the construction and completion of oil and gas wells will naturally diminish. Ohio will not drill 1,600 new wells each year ad infinitum. However, the energy produced from those wells is expected to be produced for several decades, and if it does, those volumes are expected to apply downward pressure on energy prices, generating significant cost-savings for downstream users, and creating high-wage, long-term jobs in the process.

According to Partridge, et al., a more preferred approach to modeling the potential economic impacts of development would entail waiting several years until more data comes in, and then preparing analyses comparing shale-producing counties to those without activity. As a general proposition, we agree. But unfortunately, we do not have access to a crystal ball. In absence of the ability to collect data that has not yet been generated, we apply a standard input-output methodology to characterize the potential for jobs and inter-industry purchases in Ohio. Armed with these forecasts, Ohioans can be trained to work in and with this industry, and share in its growth. And investors can make better, more-informed decisions regarding their investments with forward looking estimates.

Specific to the Partridge paper, the authors use what they admit is a rule of thumb multiplier of two and assume that direct expenditures are only found in the oil and gas / mining sector. (p. 12)  In fact, the Considine study (and ours, indirectly) uses a detailed expenditure budget that allows us to appropriately spread investments by the industry across a variety of industries. So, in fact, we (and Considine) also account for direct expenditures in many other industries.  Moreover, the use of the REMI model is far different than the use of IMPLAN or RIMS multipliers.  The REMI model is a dynamic forecasting and policy analysis tool that incorporates the complete inter-industry relationships found in input-output models. It, however, integrates input-output, computable general equilibrium, econometric, and economic geography methodologies. It takes into account interregional inter-industrial connections; commuting information, job relocation and population changes.

Considine uses a multiplier of 2.07, well within range of the 2.0 multiplier used by Partridge, et al. A simple calculation of direct expenditures yields at least 100,000 jobs. Assuming each well costs $10 million to site, construct, drill, complete and prepare for production – we multiply that figure by 1,650 wells and then divide it by $150,000 output per employee. Hence our estimate of 110,000 jobs.

Regarding Partridge’s assertion that a multiplier of two is on the high end, the Federal Reserve System’s Fiscal Impact User Guide states:

[N]ational output multipliers tend to range between 2.5 and 4.0 depending on the industry.

That is, it is assumed that each $1.00 of direct activity spurs $1.50 to $3.00 of indirect and induced activity (be it output, wages, income, or employment).

Partridge asserts that we used the “old” input output method that does not account for what would have happened without the energy development.  (footnote, p 5, and p.11, paragraph 3).  Of course, in our study, we explain that using REMI, we report the difference between the baseline growth (what would occur if the Shale investment would not take place) and the stimulated growth (p. 11 of our report).

In response to Partridge’s statement that “Foremost, impact studies are not viewed as best practice by academic economists and would be rarely used in peer reviewed studies by urban and regional economists.” There is an entire professional association devoted to the study and refinement of these methods called the International Input-Output Association. It publishes a double-peer reviewed journal entitled Journal of the International Input-Output Association and its editorial board includes a member of Ohio State faculty.  Moreover, the Journal of Regional Science recently reviewed an update of a classic input-output text by Miller and Blair (Volume 51, Issue 1, pages 196–197, February 2011).  The reviewer notes:

“While selected material in the opening chapters will be useful for undergraduate teaching, we expect the extensions in the second part of the book will be most useful for postgraduates, researchers, and academics.”

Partridge is an editor for the Journal of Regional Science.

Partridge additionally indicates that we did not adequately explain the meaning of “created and supported” when we write of jobs (footnote 2, page 5.).  We do offer an explanation of employment, but not specifically created and supported.  We define employment in terms of jobs. This includes full-time, part time, and temporary positions. A job is equal to the annual average of monthly jobs in that industry (the same definition used by QCEW, BLS, and BEA nationally). Thus, one job lasting 12 months is equal to two jobs lasting six months each, which in turn is equal to three jobs lasting four months each. A job can be either full-time or part-time.

Partridge, however, conveniently uses the word “jobs” throughout the paper and refers to BLS job growth over a six-year period when inaccurately calculating his rule of thumb multiplier (p. 12, Partridge report). His report uses employment in two industrial sectors and ignores other industries that are involved in direct spending associated with shale-related drilling and exploration.

In short, although our study benefited from insight, data and funding provided by industry, it is a study that is sound in its methods and reasonable in its conclusions. Our firm could not have built the reputation that it has over the past 15 years if we were to compromise on this approach. You can be assured that in this case, as in all others, we did not.