New Developments in Social and Behavioral Research: The Social Observatories Coordinating Network, September 2016
The nation and the world are changing rapidly, and yet we as social scientists are trying to understand what is happening with tools and infrastructure that have not changed much since the 1960’s. To make this situation even more challenging, new social phenomena and particularly social media have arisen that cannot be studied with traditional methods, but our academic programs do not prepare the new generation of social scientists to link new media with other kinds of data.
Developments in information technology offer an unprecedented opportunity to collect diverse data at fine-grained spatial and temporal scales, to change the way social, behavioral and economic science (SBE) is conducted, and to greatly expand the questions that can be addressed. Today, the proliferation of new data coming from the internet and social media require new ways to link across social science disciplines and to link social science with genetic, linguistic, medical, environmental, biological, and earth systems science. This is an opportune time to rethink the primary ways in which social and behavioral data are collected, gathered, coded, curated, documented, archived, and disseminated in the U.S. A series of workshops sponsored by NSF (in 2005, 2007, 2008, 2009, 2010, and 2011) led to a consensus that, working together across-disciplines, the social and behavioral science community could address these challenges. In 2012 the National Science Foundation began to support a set of researchers and academic faculty from across the U.S., aka “the Social Observatories Coordinating Network “(SOCN), to take on this challenge. In our several years of discussions (see materials at www.socialobservatories.org) the network settled on a challenge for the future-- to design a national network of regional data centers that could be coordinated through common objectives, sharing of protocols, and data sharing (Moran et al. 2014).
The network concluded that behavior is so situation- and place-specific that it is practically impossible to use un-clustered and widely dispersed national samples of populations to draw conclusions about processes in any one place. Populations tend to be spatially clustered by characteristics. This has implications for important issues of national concern. For example, the U.S. has always been characterized as the land of opportunity, where anyone can, through effort, succeed and attain the American dream. Yet, a recent study shows that 1 out of 4 children raised in the middle class has slipped downward by their early 40’s (Acs, 2011)-- even more important is the finding that the chances for upward mobility and its maintenance depend importantly on where people live (Olinsky and Post, 2013; Chetty et al., 2013). Similarly, a University of Washington study revealed that life expectancy for American males (and females) varies by up to 18 years, depending on which American county they live in (Institute for Health Metrics and Evaluation, 2103). That is a remarkable range, and social scientists are just now mapping the social indicators to document it and put it in context. Research focusing on context has the promise of pointing to pockets of concentrated disadvantage and poor health, where resources can be targeted to do the most good.
Based on our discussions with a wide range of scientists, our network has proposed the development of a set of 20 to 25 regional data centers located across the United States in which each center would seek to collect, organize, create, and disseminate data. These regional data centers or “social observatories” would follow 400 census tracts over time and space from these 20-25 regions spread across the country. Working closely with local and state governments they will access administrative data that will provide not a sample of the population of those several hundred census tracts, but complete records on all the population in those tracts (i.e. circa two million Americans). The centers would serve as data collection facilities wherein data are cleaned, linked, and made available for legitimate research purposes through a highly secure integrated data dissemination system. Although they would be charged with keeping data on their particular geographic region, some of these centers may have a national focus as well. They may conduct surveys but they would also use data sources that until now have not been part of the toolbox of the social sciences, and connect these data to local context and place without losing a capacity to aggregate and serve as a national sample of people and places in the USA. Our vision is that, collectively, they would offer a nationally representative sample, one that is highly clustered so as to capture local context and variability.
The major substantive foci of the data centers would be initially on questions of a) change and adaptation and b) opportunity and mobility, both of them broad questions that require data linkages and granularity in data sources. Because they would focus on place and context as well as people they would make it possible to determine which kinds of investments in infrastructure have the most impact on well-being and how economic change can be fashioned to provide the greatest opportunities with the fewest barriers to mobility to its people. They could also provide information on the organizational structures of communities that may facilitate or hinder appropriate adaptation to ongoing economic, social, and environmental change.
Having these networked regional data centers will transform how the SBE sciences go about their work by encouraging the integration of the SBE sciences, rather than promoting the fragmentation that we have experienced since the 1960’s. The latter was a necessary phase to achieve greater specialization but has had over the years the effect of making it ever more difficult for SBE scientists to share methods and approaches to address issues of national importance. The regional data center network will explicitly promote what is now a broad call from the National Academy of Sciences to integrate the social and physical sciences to address issues of importance with the best tools available without regard for disciplinary origins. Along these lines, a consortium of scientists led by Michael Barton of Arizona State University has been developing what could become a National Center for Human Systems Science. This Center would utilize insights from the behavioral sciences along with those from the physical sciences to improve our understanding of human systems and help design solutions to ecological problems of local and global scales including natural and human-caused hazards.
The network is preparing an edited volume of papers from a March 2016 conference that will be published in the Annals of the American Academy of Political and Social Science (“New Developments in Data Collection: Linking Data across Levels”) in early 2017. For more information about this project and a list of network members please go to the website Socialobservatories.org or contact PI Sandra Hofferth, University of Maryland (firstname.lastname@example.org), or co-PI Emilio Moran, Michigan State University (email@example.com).
Acs, G. 2011. Downward mobility from the middle class: Waking up from the American dream. Pew Charitable Trusts, Washington, DC.
Chetty, R., N. Hendren, P. Kline, E. Saez, 2013. The economic impacts of tax expenditures; Evidence from spatial variation across the U.S. Equality of Opportunity Project, Harvard University and the University of California, Berkeley, July 2013. http://www.equality-of-opportunity.org/
Institute for Health Metrics and Evaluation. 2013. The State of US Health: Innovations, Insights, and Recommendations from the Global Burden of Disease Study. Seattle, WA: IHME.
Moran, E.F., et al. 2014. Building a 21st century infrastructure for the social sciences. Proceedings of the National Academy of Sciences, November 11, vol. 111, no. 45 , 15855–15856. (www.pnas.org/cgi/doi/10.1073/pnas.1416561111).