Scientists ponder the politics of data

BOSTON'Few government policy decisions are more fraught with political controversy than those involving land use'whether, say, forest policy at the federal level or zoning at the local.

Yet the computer simulation models many policy makers use to predict the outcomes of proposals aren't up to the task of giving objective answers.

Two researchers at the University of Washington are trying to change that. Alan Borning and Paul Waddell, co-directors of the Center for Urban Simulation and Policy Analysis, have been working on an urban simulation model that is sensitive to political questions. They applied an early version of UrbanSim to Honolulu in 1996; the latest version is being used by Salt Lake City officials.

They outlined their progress at the Conference on Digital Government Research, sponsored by the National Science Foundation.

'You start out with policies of interest, and see how those policies would affect outcomes, then make your [simulation] algorithms sensitive to those policies,' Borning said. For example, many municipalities and counties want to know what effect on driving and walking patterns a certain housing development might have. In nearly every part of the country, Waddell pointed out, traffic congestion and other growth-related quality-of-life issues are hot topics.

Existing transportation models, he said, are too coarse geographically to really answer such questions with any accuracy. He and Borning are using a variety of techniques to improve the accuracy of simulations, but much of their work with the Wasatch Front Regional Council in Salt Lake City involves painstaking loading of historical inputs and trying to build the math to recreate known outputs automatically.

'Model validation is a major issue,' Waddell said. Even a valid model for some variables can't predict extraordinary events such as the siting of a major mall or the failure of a large employer.

Other researchers at the conference said that even with objective models, political considerations often color policy decisions, no matter what the data show.

Said Niki Steckler of the OGI School of Science and Engineering at the Oregon Health and Science University, 'Politics drives data much more than data drives policies.' But, she added, scientists need to understand that public policy decisions are almost never totally rational, based purely on data. Yet good policy-making can't take place in the absence of objective data.

'Neither is a productive way to manage public resources,'
Steckler said.

inside gcn

  • artificial intelligence (ktsdesign/Shutterstock.com)

    Machine learning with limited data

Reader Comments

Please post your comments here. Comments are moderated, so they may not appear immediately after submitting. We will not post comments that we consider abusive or off-topic.

Please type the letters/numbers you see above

More from 1105 Public Sector Media Group