Carbon efficiency of US colleges and universities: a nonparametric assessment

ABSTRACT

Under the American College and University Presidents’ Climate Commitment (ACUPCC), institutes of higher education have pledged to pursue a goal of carbon neutrality. We utilize emissions reported under the ACUPCC agreement and a nonparametric data envelopment analysis approach in order to evaluate the relative performance of signatories to the agreement in terms of producing teaching and research with the least greenhouse gas emissions. We find that while many signatory institutions are now producing their desirable outputs relatively efficiently in terms of carbon emissions, there still exists considerable variation in efficiency and potential for improvement. Results of a second stage efficiency change analysis shows evidence of both movement towards the efficiency frontier since signing, and some movement of the frontier itself, though this evidence comes primarily from teaching-focused institutions.

KEYWORDS:

• ACUPCC
• data envelopment analysis (DEA)
• range-adjusted measure
• undesirable outputs

JEL CLASSIFICATION:

• Q56
• C14

Acknowledgements

The authors thank Claire Donohue and Sara DiMassimo for research assistance, and the Colgate University Research Council for a grant for submission fees.

Disclosure statement

No potential conflict of interest was reported by the authors.

Supplementary material

Supplemental data for this article can be accessed here.

Notes

¹ Now housed at Second Nature (http://secondnature.org/)

² As of July 2015, there were 685 signatories to the ACUPCC.

³ The use of the ranges as weights on the slacks in the objective function serve to normalize the slacks in the sense of cancelling the units and making them units invariant. This can be achieved by dividing by other first order dispersion measures. Lovell and Pastor (1995) use the standard deviation, but suggest that the range is another option. Cooper, Park, and Pastor (1999) fully develop the version using the range.

⁴ In the presence of multiple outputs, there are alternatives to using DEA. For instance, by using a translog output distance function and assuming linear homogeneity of outputs, a stochastic distance function approach could be applied. (See, for example, Grosskopf et al. (1997) and Coelli and Perelman (1999).)

⁵ http://rs.acupcc.org/

⁶ Of course, it is possible that university systems, such as the University of California (UC), may wish to rank the relative environmental performance of institutions within the system. For instance, UC administrators may be interested in efficiency variations between UC San Diego and UC Berkeley, two peer institutions within the UC system. For this case, the approaches offered by Azad and Ancev (2010, 2014) provide future research possibilities.

⁷ Thrall (1996) examines outcomes based on various subjective criteria for the weights.

⁸ There is no correlation between the 2011 efficiency scores of institutions who joined the agreement in 2010–2011 and those who joined earlier. One explanation for this is that institutions joining in later periods were already taking steps to reduce their emissions and joined ACUPCC to advertise this.

⁹ http://nces.ed.gov/ipeds/datacenter/

¹⁰ http://mup.asu.edu/

¹¹ http://www.leidenranking.com/

¹² http://www7.ncdc.noaa.gov/CDO/CDODivisionalSelect.jsp

¹³ https://www.commonfund.org/CommonfundInstitute/HEPI/Pages/default.aspx

¹⁴ See the Interagency Working Group on Social Cost of Carbon, United States Government: https://www.whitehouse.gov/sites/default/files/omb/inforeg/scc-tsd-final-july-2015.pdf.