1. Brendon Swedlow Department of Political Science Institute for the Study of the Environment, Sustainability, and Energy Northern Illinois University Comparative Risk Regulation Workshop University of California, Berkeley December 13-14, 2012

2.  Case Studies  Quantitative Analysis of Existing Indicators

3.  Case Studies  Lack External Validity  Limited number  Quantitative Analysis of Existing Indicators  Lack Internal Validity

4.  Mixed Methods  Comparative Nested Analysis of Cases Lieberman APSR 2005 Swedlow, Kall, Zhou, Hammitt & Wiener Law & Policy 2009

5.  Representative Sample of Cases  Research Questions of Wide Interest  Distributed Data-gathering Approach  Data-gathering integrated into course instruction  Comparative Nested Analysis of Cases

6. 11,086 “verbatim” risks from 254 sources in literature on risk perceptions, ranking, and classification, 1960-2003, in US and Europe Universe of all risks 3,000 “unique” risks (recom- bining essentially identical “verbatim” risks) In 19 categories and 95 sub- categories 2,878 “unique” risks In 18 categories and 92 sub- categories (dropping 122 unique risks, 1 category, and 3 subcategor ies) 100 in random sample; 92 in stratified random sample Constructing a Representative Sample

7. Table I. Risks by Type Percentage in: Code Category Matrix Sample 1 Crime and violence 1.8 3 2 Alcohol, tobacco, and other drugs 3.0 3 3 Medication and medical treatment 6.8 8 4 Transportation 8.2 13 5 Accidents not elsewhere classified 2.4 2 6 Recreation 5.5 8 7 War, security, and terrorism 1.5 3 8Toxic substances 9.8 8 9 Food and agriculture 9.5 9 10 Pollution 7.5 8 11 Energy production 5.0 3 12 Political, social, and financial 3.4 1 13 Ecogeological 4.0 2 14 Global 2.2 1 15 Human disease/health 9.7 9 16 Occupational 15.017 17 Consumer products 3.4 2 18 Construction 1.4 0 Total percentage 100 100 Total number 2,878 100

8. 100 Risks Randomly Selected from a Universe of 2878 Risks (and their distribution across 18 Risk Categories and 92 Risk Subcategories) Risk Categories (18) and Subcategories (92) 100 Representative Risks 1. CRIME AND VIOLENCE RISKSBurglary Sabotage Firearms 2.ALCOHOL, TOBACCO, AND OTHER DRUG RISKS Alcohol risksSake Tobacco risksSmoking regulations Drug abuse (Non-prescription illegal drugs) risks Pot smoking 3.MEDICATIONS AND MEDICAL TREATMENT RISKS Adverse drug reaction risksShortage of medicines Caffeine - chronic effects Medical procedure risksChildbearing Radiation therapy Medical error risks not elsewhere classified Vaccination risksVaccination - side effects X-ray risks Other medical risksHealth care facilities and services - exposure to physical agents Health care facilities and services Genetic engineering

9.  Theories of regulation are propositions or sets of hypotheses about  Why regulation emerges  Which actors contribute to that emergence  Typical patterns of interaction among regulatory actors -- Morgan & Yeung 2007  All theories of regulation will have to explain variation in  Information gathering  Standard setting  Behavior modification -- Hood, Rothstein, & Baldwin 2001

10. Research Questions of Wide Interest (1) How is the risk defined and assessed? (2) Who regulates the risk? (3) How is the risk regulated? (4) Why is the risk assessed and regulated the way that it is? (5) What are the consequences of assessing and regulating the risk the way that it is?

11.  Collaborative Research Network (CRN)  Relies on Researchers Collecting Data in Countries where they are based  Advantages:  Uses native knowledge  Cheaper  Can be sustained over time

12.  Initial data-gathering done by students in research paper assignments  First three research questions have been answered for US & Illinois for almost all 100 risks  Now focusing on sub-sample of 30 risks  Martin Lodge’s LSE students now studying same 30 risks in UK

13.  Size of Risk Equals Size of Regulatory Effort  If Size of Risk Not Equal to Size of Regulatory Effort, then Regulatory Effort explained by  Market Failure  Public Opinion  Interest Groups  Political Culture  Etc. -- Hood, Rothstein & Baldwin 2001

14.  RAs extracting information from student papers on  Size of risk  Size of regulatory effort  RAs will code size of risk & regulatory effort as large, medium, or small  Coding will allow regression analysis of how size of risk relates to size of regulatory effort  Regression analysis will reveal cases where these are in alignment and cases where they are not  Both kinds of cases will be studied further to determine causes of differences in risk assessment & regulation