Statistics are not religion

June 4, 2012
Peter Garforth says statistical probabilities around climate change are merely guides for making energy-related decisions.

In the majority of cases, sound energy management practices are more than justified by their reliability, costs and risk-mitigation value to a business. Significant competitive advantages accrue to companies with a sustained focus on energy productivity. Generally these same companies also achieve significant reductions in energy-related greenhouse gas emissions, commonly called the carbon footprint.

Rather than just treat emission reductions as a desirable outcome of economic and reliability benefits, this aspect should be addressed as a goal in its own right. In recent years, the climate change debate in the United States has become so polarized between “believers” and “non-believers” that it sometimes feels more like a discussion of religious beliefs rather than an assessment of scientific data. This fault line has more recently taken on a political flavor as evidenced by the changing positions of members of all political parties.

View more content on


Depending on personal affiliations, perceptions and sources of information, this schism may be reflected in how management assigns priorities and investments when carbon footprint reduction discussions are on the table. Investments with questionable economics but a significant carbon reduction may be implemented with damaging effects on the business. Alternatively, opportunities to incrementally reduce carbon with some lowering of immediate return economics during the implementation of efficiency measures or other investments may be missed.

It is probably time to revisit the basic question around energy and climate change. As a general rule, the average temperature of the earth is highly correlated to the concentration of greenhouse gases in the atmosphere, whether they come from human or natural sources. There is extensive data going back hundreds of thousands of years and thankfully little dispute over this basic evidence. The debate is whether human use of oil, coal, and gas for energy in the last 200 years is a major contributor to a new imbalance that is accelerating the warming of the atmosphere and creating more frequent and more intense weather events.

The statistical probability that can be explained by human activity is in the 90th percentile from the available data. As with all probability assessments, there is the possibility that some other hypothesis explains the data, with naturally occurring variations unrelated to humanity being the most commonly cited. The debate should be between the probabilities of different hypotheses, not between personal or political belief systems.

Peter Garforth heads a specialist consultancy based in Toledo, Ohio and Brussels, Belgium. He advises major companies, cities, communities, property developers and policy makers on developing competitive approaches that reduce the economic and environmental impact of energy use. Peter has long been interested in energy productivity as a profitable business opportunity and has a considerable track record establishing successful businesses and programs in the US, Canada, Western and Eastern Europe, Indonesia, India, Brazil and China. Peter is a published author, has been a traveling professor at the University of Indiana at Purdue, and is well connected in the energy productivity business sector and regulatory community around the world. He can be reached at [email protected].
Subscribe to the Energy Expert RSS feed

Decisions by businesses at a minimum should assess and quantify the risks associated with the higher and lower probability scenarios. Teams must feel comfortable including these risks as a part of the overall management discussions without any spoken or unspoken “no go” areas. Organizations and companies that have a general acceptance of the lower probability hypothesis would generally assign a low priority to carbon footprint reduction. In terms of the energy programs, they would place a high value on efficiency, and a relatively low value on switching to low-carbon sources of fuel, unless they also happened to be substantially cheaper than the present options. This would include switching to on-site clean and renewable supplies and “green power” procurement.

This view would also assume that fuel and electricity pricing would not be significantly impacted in the future, depending on their carbon content. The assumption would be that carbon pricing and carbon regulation would have minimal market price impacts. In some case, this may even relate to the extraction process itself, as could be the case with shale gas and oil.

Last, but not least, managing on the basis of the low probability hypothesis would not anticipate any significant deterioration in energy supply reliability or production continuity as a result of intensified or more frequent weather events. This will generally bias against investments to harden on-site or near-site supply systems, or to increase the resilience of the site in general to weather risks.

Companies investing on the basis of the high-probability hypothesis will paradoxically do most of the same things around energy efficiency. The key difference will be the future value they will attach to carbon reduction caused by efficiency. As a result they may accept a somewhat lower immediate return to mitigate a higher future risk. However, this lower return should generally never fall below minimum accepted corporate hurdle rates.

In a similar way, these companies will consider reconfiguring energy supply both in terms of on-site options and procurement approaches based on the future value of carbon. They would also evaluate these based on an accelerated reduction in equipment costs based on the assumption of higher volume markets and innovation. Again, slightly lower immediate returns could be the case, but still well within acceptable corporate limits. The possible value of on-site supply and other reliability investments would be assessed not only on their energy values, but also on the avoided costs of production interruption.

Neither of these approaches is right or wrong. They simply reflect differing views of the probability and effects of managing carbon-based emissions from fuel. It is probably time for us all to remember that the statistical probabilities around climate change are not religion but helpful guides to making difficult energy-related investment decisions.