The effect of administrative form and stability on cities’ use of greenhouse gas emissions inventories as a basis for mitigation

ABSTRACT

Sparked by their emergence as innovative climate policy leaders, cities’ decisions to engage in greenhouse gas (GHG) mitigation planning have been the subject of extensive examination over the last decade. The impact that these planning and subsequent local policy actions have on addressing actual emissions, however, remains under-explored and even less is known about the conditions under which cities’ climate planning efforts result in meaningful mitigation. Data limitations, specifically the sporadic and unstandardized accounting of local GHG emissions, underlie the minimal empirical attention given to understanding cities’ climate policy outcomes. To circumvent these data challenges, we quantify the impact of local efforts through expert evaluation. We utilize nation-wide survey data collected from U.S. cities in 2010 and 2015 to empirically assess how community and city government characteristics influence the extent to which cities’ climate planning efforts serve as the basis for emissions reductions. The results suggest that the dynamics shaping cities’ decisions to complete an inventory are different from those that influence whether it is used as a basis for emissions reductions. Results also point to the positive effect of regular inventory updates and to the importance of stability in the administrative leadership of sustainability in a city’s government.

KEYWORDS:

• Local climate protection
• cities
• greenhouse gas
• emissions inventories

Disclosure statement

No potential conflict of interest was reported by the authors.

Notes on contributors

Rachel M. Krause is associate professor in the School of Public Affairs and Administration at the University of Kansas and director of its Masters of Public Administration (MPA) program. Her research focuses on local governance, urban sustainability policy, and municipal climate protection initiatives.

Angela Y. S. Park is assistant professor in the Department of Political Science at Kansas State University. Her research focuses on understanding the key challenges facing local governments in delivering sustainability services and programs and what enables them to overcome these challenges. She is particularly interested in the effects of institutional arrangements in dealing with the issues of interagency coordination and performance management.

Christopher V. Hawkins is associate professor in the School of Public Administration at the University of Central Florida. His research focuses on urban politics, metropolitan governance, and urban sustainability policy.

Richard C. Feiock is the Augustus B. Turnbull Professor of Public Administration and holds the Jerry Collins Endowed Chair at Florida State University. He researches local government, sustainability, and local democratic institutions. He is a National Academy of Public Administration fellow and served on the Board of Scientific Counselors for the U.S. Environmental Protection Agency.

Notes

1 The ICSD is the source of 3 of the independent variables in the model. Rather than using listwise deletion, which would drop all observations missing any one of these values, leading to a loss of power and causing potential bias, we utilize multiply imputed data and appropriate estimation techniques. As Curley et al. (2017) describe in detail, the imputation approach employed by the ICSD identifies a theoretically informed and statistically supported set of predictor variables for each target variable, which are then used to generate 20 rounds of estimates for each value. When utilized with appropriate estimation techniques, this ensures that the uncertainty inherent in the prediction of missing values is accounted for with increased standard errors (Rubin, 1987; van Buuren, 2012).

2 Normally, we would utilize a selection model, specifically in this case a Heckman Ordered Probit, to account for the possibility that a city’s decision to conduct an inventory is not independent from its subsequent impact. However, this procedure is not designed to run on imputed data. When the estimation was ‘forced’ in Stata, the models fail to converge. Instead we treat these two dynamics as separate from each other and are careful to reflect that in our interpretation and discussion of results.

Additional information

Funding

This work was supported by the National Science Foundation [grant number 1461526/1461506/1461460]. Any opinions, findings, and conclusions expressed are those of the authors and do not necessarily reflect the veiws of the National Science Foundation.