Social Observation for Sustainability Science
Social Observation for Sustainability Science[1]
Contribution to the
Social Observatory Coordinating Network
John B. Braden
Professor Emeritus of Environmental Economics
University of Illinois
Department of Agricultural & Consumer Economics
1301 W. Gregory Drive, Rm 326
Urbana, IL 61801
February 2013
Social Observation for Sustainability Science
1. Introduction/Abstract
Sustainability Science is an important new intellectual frontier [Clark 2007; Kates et al. 2001]. I contend that (1) coupled observation of humans and nature at large scale is important to many fundamental questions of sustainability, (2) current programs observing natural and social systems are inadequate to address fundamental questions of sustainability at large scale and (3) social scientists and natural scientists/engineers agree on many characteristics of coupled data needed to advance sustainability science.
2. Data for Sustainability Science
Sustainability science studies interactions between human and natural systems. These interactions are interesting on their own and critical to harmonizing economic and social development with resource availability and environmental quality. This new field challenges the scientific community to collect data that “encompass[es] different magnitudes of scales (of time, space, and function), multiple balances (dynamics), multiple actors (interests) and multiple failures (systemic faults)" [Reitan, 2005]. Braden et al. [2010] call for “data [collected] across time and space that enable evidence on perceptions, attitudes, social institutions, situation-behavior relationships, and decision-making to be linked comprehensively to measurements of the natural environment.”
Due to the complexity of human-nature interactions, most sustainability science has been conducted at small scale through localized observation and analysis. However, many important questions require research at large scales where the effects of climate, terrain, preferences, culture, legal and regulatory systems, information sources, and other spatially and temporally variable factors can be examined. Example include: 1) how attitudes, preferences and environmental quality interact to influence the evolution of regulatory institutions, and 2) how culture environmental conditions interact to influence human response to information about the environment.
3. Illustrating Flaws in Current Observation Programs
Current data collection programs provide little support for coupled research at large scale. Two examples powerfully illustrate the problem [Braden, Jolejole-Foreman and Schneider, 2013b]. The examples investigate whether measured water quality affects behavior – respectively, recreational choices and household water conservation efforts.
In the US, national scale provides considerable variation in environmental conditions and local cultures. USEPA’s National Aquatic Resource Survey program (http://water.epa.gov/ type/watersheds/monitoring/aquaticsurvey_index.cfm0, Feb. 13, 2013) provides point observations on nearly 200 water quality metrics. Its sampling frame is statistically representative of nine US ecoregions. The decadal environmental module of the NSF-sponsored General Social Survey (GSS) (http://www.norc.org/Research/Topics/Pages/ environment.aspx, Feb. 13, 2013) provides data on selected environmental attitudes and water conservation efforts. Its sample of 2,044 households is statistically representative at national scale. Finally, the US Forest Service-sponsored National Survey on Recreation and the Environment (NSRE) (http://warnell.forestry.uga.edu/nrrt/nsre/clemsc907.html , Feb. 13, 2013) provides detailed recreation information for a nationally representative sample of 8,073 households.
Privacy concerns limit the availability of GSS and NSRE data to census tract (GSS) or zip code (NSRE) aggregates. Using stream water quality indicators collected by EPA at 2,042 points, only 16 GSS observations are collocated with the water quality observations. Further aggregation to county level achieves 195 collocated observations, sufficient for statistical analysis, while also reducing variation through averaging. For recreational observations, 217 NSRE observations are collocated at zip code-level with stream water quality data. This sample size supports statistical analysis. Thus, the geospatial matching process eliminates nearly 90% of the original information. Importantly, comparing the means of the raw and matched samples for 46 variables used in estimating the two relationships reveals statistically significant differences in all but four cases. Thus, the matched samples are almost certainly unrepresentative of the underlying populations—both human and natural—at any scale.
Using the matched samples, significant relationships are apparent between water quality, attitudes, behaviors and, notably, regional differences, particularly for recreation. The significance of regional differences reinforces the need for analysis at large scale. In principle, the findings could be important for designing information programs about recreational choices and water conservation programs. But, the unrepresentative matched samples make generalization of the findings impossible. The only solution is a data collection program designed from the outset to be jointly representative of human and natural systems.
4. Disciplinary Data Aspirations
A multi-disciplinary convenience sample of water scientists and professionals surveyed by Braden et al. [2013b] reveals strong interest in addressing human dimensions of water science, even among natural scientists and engineers (NSE). Agreement is apparent across disciplines that coupled data are most needed at the household and community scales, although social scientists (SS) and NSE differ over their relative priority. SS seek panel data while NSE seek an exhaustive inventory of natural parameters. Finally, the disciplinary differences in data aspirations are greater for questions with strong policy implications than for science-oriented questions.
5. References
Braden, J, C Jolejole-Foreman, D Schneider 2013a, Humans and the water environment: the need for coordinated data collection, Univ. Illinois, Urbana, in review.
Braden, J, D Brown, D Maidment, S Marquart-Pyatt 2013b, Populating the water world, draft, Univ. Illinois, Urbana.
Braden, J, et al 2010, Coupling human system data with natural system data: laying a foundation for sustainability science, Sub. to Soc., Behav. Econ. Sci. Div., Nat. Sci. Fndn.
Clark, WC 2007, Sustainability science: a room of its own, Proc. Nat. Acad. Sci. 104: 1737-1738.
Kates, R. et al 2001, Sustainability science, Science 292(5517): 641–642.
Reitan, P. 2005, Sustainability science – and what’s needed beyond science, Sustainability: Science, Practice, & Policy 1(1):77-80.
[1] The research reflected here has been supported, in part, by awards CBET-0838607, CBET-0827497, and EAR 1038813 WSC of the National Science Foundation and USDA/National Institute for Food and Agriculture project ILLU-470-316 of the Illinois Agricultural Experiment Station. The author alone is responsible for all aspects of the analysis, interpretation, and conclusions in this paper.
