2.2. Identifying Potential Climate Compatible Development Efforts and Missing Links
For each of the sixteen policies in each of the 287 responding cities, I first assessed whether or not the city had undertaken the policy effort and said that considerations about economic development, sustainability, climate change mitigation, and climate change adaptation had all influenced it. Because researchers have struggled using formalized means of studying nascent climate change policy efforts in North American cities [38
] and because climate change compatible development efforts are grounded in policymakers deliberatively balancing various contextually-based objectives; I decided to allow respondents to identity which of these objectives had influenced their city government’s considerations.
If the respondent identified that a policy action was influenced by considerations about economic development, sustainability, mitigation, and adaptation; I considered that an instance of potential climate compatible development in the city. In order to gain more perspective on the presence of potential climate compatible development in these cities (or not), I also explored instances in which each of these issues acted as “missing links”—when they were the only one of the four issues not influencing a policy action the city was undertaking. For example, if a respondent described that considerations about sustainability, mitigation, and adaptation influenced their city’s efforts related to increasing pedestrian transportation; but that economic development did not, then economic development was acting as a missing link for that policy action.
2.3. Regression Analysis
In order to explore contextual factors that might be related to the potential for climate compatible development efforts, I also performed two logistic regression analyses. The first was a binary logistic regression that tested the association between independent variables described in the next paragraph and a dependent variable describing whether (1) or not (0) a city had at least one potential climate compatible development effort.
I chose the independent variables used in these models based on inspiration from Ellis et al.’s [25
] list of drivers and challenges for climate compatible development and adapted them for the context being studied. Table 1
summarizes Ellis et al.’s list along with explanations of the associated independent variables used in my analysis. Value inflation factor tests performed on the models found no evidence of significant multicollinearity. In their paper, Ellis et al. [25
] first describe recognition of the need for adaptation to bolster resilience, foster growth, and reduce poverty. Those living in the Great Lakes region have spent decades in collaborative efforts to clean up their unique freshwater environment while attempting to re-orient their development strategies towards taking advantage of amenities associated with freshwater resources [34
]. These efforts have shaped the development of climate change adaptation efforts in the region [40
]. Therefore, the first variable included in the model was the city’s “water footprint”—the area of the city composed of water versus land. Presumably, in this region, the larger the city’s water footprint the more likely it would be to pursue climate compatible development. Ellis et al. [25
] also describe a perceived need for energy or resource efficiency. With that in mind, I included a measure of the city’s budget surplus versus shortfall 2005–2010. Cities in the US are under substantial pressure to balance their budgets each year; and budget deficits have been associated with more climate change mitigation efforts [11
]. Therefore, cities that are experiencing fiscal stress in the form of expenditures recently exceeding revenue might be more likely to connect their climate change efforts with meeting their financial concerns. Next, Ellis et al. [25
] describe the desire to take advantage of new economic opportunities as a driver of climate compatible development. Therefore, I included a measure of the city’s poverty rate as a higher poverty rate should create more pressure to find new economic development strategies. Based on Ellis et al.’s [25
] description of a desire to access aid, I then included the percentage of the city’s revenue composed of funds from higher levels of government from 2005–2010 (intergovernmental funds), with the presumption that higher dependence would be associated with more potential climate compatible development. Ellis et al.’s [25
] final driver was strong government leadership. This inspired me to include a binary variable describing whether (1) or not (0) a sustainability or climate change policy entrepreneur were present in the city based on my survey results (more information about the development of this variable can be found in a related study [35
]). Studies have demonstrated that the presence of environmental policy entrepreneurs are not only critical in the emergence of climate change policy efforts [41
], but in making connections between climate change efforts and other issues [1
Ellis et al. [25
] also include six challenges to the emergence of climate compatible efforts: costs, opposed interest groups, a lack of awareness or trusted information, short-termism, lack of capacity to respond to and implement strategies, and institutional constraints. For costs, I included a measure of the city’s debt burden—how much of its expenditures were devoted to servicing existing debts 2005–2010. Cities in the US compared to others throughout the world have a relatively large amount of freedom to take on debt to finance their investments; however, because they face pressure to balance their budgets year to year, having to pay off existing debts can constrain taking on new policy initiatives. Therefore, I hypothesized that the higher the city’s existing debt burden, the less likely it was to perceive that it had the flexibility in its budget to take on potential new costs related to innovative efforts related to climate compatible development. For opposition, I included a variable describing the percentage of the vote Mitt Romney (R) received versus Barack Obama (D) in the 2012 presidential election in precincts in each city. Climate change remains a divisive issue in the US, and political party support remains a strong predictor of potential opposition to developing climate change-related policies [41
]. For awareness and trusted information, I included a measure of the bachelor’s attainment rate in the city. Presumably the percentage of city residents that had attained at least a four-year college degree should be positively related to greater levels of awareness and trust in information about climate change in the city population that would provide a base of support for city politicians and city staff to have opportunities to engage with scientific research and feel more comfortable forging connections between policy efforts and science [41
]. For short-term outlooks limiting long term planning, I included the percentage of the population employed in the real estate sector in the city. I hypothesized that a greater number of residents employed in this sector might lead to considerations about city development having longer time horizons. To address the capacity to respond to and implement strategies, I included a binary variable describing whether (1) or not (0) the city had a Council-manager form of government. Council-manager forms of government are a popular local government reform strategy in the US that are believed to make government more efficient and have been associated with cities taking on climate change and sustainability efforts [45
]. Therefore, I hypothesized that this form of government would be associated with climate compatible development efforts. Finally, to represent existing ability to overcome institutional constraints, I included a binary variable describing whether (1) or not (0) the city had a department or commission (groups that meet with city council legislators) that were specifically tasked with addressing environmental issues or sustainability. Such a focused area for discussion of environmental efforts would presumably be associated with more climate compatible development efforts because they could help assess and coordinate activities. In addition to these variables, I also included the population of the city (based on the 2010 Census) and the state each of the cities were in to account for variation based on the size of the city and the influence of state-level government.
To explore associations between these independent variables and particular issues acting as missing links in potential climate compatible development, I also used them in a multinomial logistic regression model. The dependent variable of this model described whether the city falls into one of four categories: there were no missing links, economic development primarily acted as the missing link, sustainability primarily acted as the missing link, or climate change mitigation or adaptation primarily acted as the missing link. To develop this variable, I first assessed what issues acted as the missing link most frequently in each city. I ignored 11 cities in which two or more issues were tied for the most frequent missing link and combined climate change mitigation and adaptation into a single category, climate change, because there were so few cities (8) where mitigation was the primary missing link. This resulted in 208 cities with no missing links, economic development primarily acting as the missing links in 17, sustainability in 19, and climate change in 32. The model compared cities in which economic development, sustainability, or climate change primarily acted as the missing link versus those cities with no missing links (i.e., “no missing links” is the reference, or baseline, category for this analysis).