Automobile Insurance: Road Safety, New Drivers, Risks, …978-1-4615-4058-8/1.pdf · AUTOMOBILE INSURANCE: Road Safety, New Drivers, Risks, Insurance Fraud and Regulation edited by
Post on 16-Feb-2018
217 Views
Preview:
Transcript
Automobile Insurance: Road Safety, New Drivers, Risks, Insurance Fraud and Regulation
CENTRE FOR RESEARCH ON TRANSPORTATION 2STH ANNIVERSARY SERIES 1971 . 1996
EQUILIBRIUM AND ADVANCED TRANSPORTATION MODELLING edited by Patrice Marcotte and Sang Nguyen
TELECOMMUNICATIONS NETWORK PLANNING edited by Brunilde Sanso and Patrick Soriano
FLEET MANAGEMENT AND LOGISTICS edited by Teodor Gabriel Crainic and Gilbert Laporte
AUTOMOBILE INSURANCE: Road Safety, New Drivers, Risks, Insurance Fraud and Regulation edited by Georges Dionne and Claire Laberge-Nadeau
TAKING STOCK OF AIR LIBERALIZATION edited by Marc Gaudry and Robert Mayes
Huebner International Series on Risk, Insurance, and Economic Security
J. David Cummins, Editor The Wharton School University of Pennsylvania Philadelphia, Pennsylvania, USA
Series Advisors:
Dr. Phelim P. Boyle University ofWaterloo, Canada
Dr. Jean Lemaire University of Pennsylvania, USA
Professor Akihiko Tsuboi Kagawa University, Japan
Dr. Richard Zeckhauser Harvard University, USA
Other books in the series:
Cummins, J. David and Derrig, Richard A: Classical Insurance Solvency Theory
Borba, Philip S. and Appel, David: Benefits, Costs, and Cyc/es in Workers' Compensation
Cummins, J. David and Derrig, Richard A: Financial Models of Insurance Solvency
Williams, C. Arthur: An International Comparison of Workers' Compensation
Cummins, J. David and Derrig, Richard A: Managing the Insolvency Risk of Insurance Companies
Dionne, Georges: Contributions to Insurance Economics Dionne, Georges and Harrington, Scott E.: Foundations of Insurance Economics
Klugman, Stuart A: Bayesian Statistics in Actuarial Science Durbin, David and Borba, Philip: Workers' Compensation Insurance:
Claim Costs, Prices and Regulation Cummins, J. David: Financial Management of Life Insurance
Companies Gustavson, Sandra G. and Harrington, Scott E.: Insurance, Risk Management, and Public Policy
Lemaire, Jean: Bonus-Malus Systems in Automobile Insurance
Automobile Insurance: Raad Safety, New Drivers,
Risks, Insurance Fraud and Regulatian
Edited by
Georges Dionne HEC, Montreal
Claire Laberge-Nadeau Universite de Montreal
.... " Springer Science+Business Media, LLC
ISBN 978-1-4613-6817-5 ISBN 978-1-4615-4058-8 (eBook)
DOI 10.1007/978-1-4615-4058-8
Library of Congress Cataloging-in-Publication Data
A C.I.P. Catalogue record for this book is available from the Library of Congress.
Copyright © 1999 by Springer Science+Business Media New York
Originally published by Kluwer Academic Publishers in 1999 Softcover reprint ofthe hardcover Ist edition 1999
AII rights reserved. No part of this publication may be reproduced, stored in a retrieval system or transmitted in any form or by any means, mechanical, photo-copying, recording, or otherwise, without the prior written permission of the publisher, Springer Science + Business Media, LLC
Printed an acid-free paper.
Contents
Guest Speakers and Contributing Authors
Referees
Preface
Preface
Foreword
Introduction Georges Dionne and Claire Laberge-Nadeau
SECTION 1 Automobile Insurance Pricing, Risks, and Asymmetric Information
1 ASYMMETRIC INFORMATION IN AUTOMOBILE INSURANCE: AN OVERVIEW Pierre-Andre Chiappori
2 EVIDENCE OF ADVERSE SELECTION IN AUTOMOBILE INSURANCE
MARKETS Georges Dionne, Christian Gourihoux and Charles Vanasse
3 ALLOWANCE FOR HIDDEN INFORMATION BY HETEROGENEOUS
MODELS AND APPLICATIONS TO INSURANCE RATING Jean Pinquet
4 BAYESIAN ANALYSIS OF ROAD ACCIDENTS: A GENERAL FRAMEWORK
ix
xi
xiii
xv
xvii
xix
13
47
FOR THE MULTINOMIAL CASE 79 Denis Bolduc and Sylvie Bonin
Contents
Guest Speakers and Contributing Authors
Referees
Preface
Preface
Foreword
Introduction Georges Dionne and Claire Laberge-Nadeau
SECTION 1 Automobile Insurance Pricing, Risks, and Asymmetric Information
1 ASYMMETRIC INFORMATION IN AUTOMOBILE INSURANCE: AN OVERVIEW Pierre-Andre Chiappori
2 EVIDENCE OF ADVERSE SELECTION IN AUTOMOBILE INSURANCE
MARKETS Georges Dionne, Christian Gourihoux and Charles Vanasse
3 ALLOWANCE FOR HIDDEN INFORMATION BY HETEROGENEOUS
MODELS AND APPLlCATIONS TO INSURANCE RATING Jean Pinquet
4 BAYESIAN ANALYSIS OF ROAD ACCIDENTS: A GENERAL FRAMEWORK
ix
xi
xiii
xv
xvii
xix
13
47
FOR THE MULTINOMIAL CASE 79 Denis Bolduc and Sylvie Bonin
viii AUTOMOBILE INSURANCE
5 COMMERCIAL VEHICLE INSURANCE: SHOULD FLEET POLICIES DIFFER
FROM SINGLE VEHICLE PLANS? 101 Claude Fluet
SECTION 2 Insurance Fraud
6 COSTLY STATE FALSIFICATION OR VERIFICATION? THEORY AND
EVIDENCE FROM BODILY INJURY LIABILITY CLAIMS Keith J. Crocker and Sharon Tennyson
7 THE FREQUENCY OF EXCESS CLAIMS FOR AUTOMOBILE PERSONAL
119
INJURIES 131 Alan F. Abrahamse and Stephen J. Carroll
8 WHEN IS THE PROPORTION OF CRIMINAL ELEMENTS IRRELEVANT?
A STUDY OF INSURANCE FRAUD WHEN INSURERS CANNOT COMMIT 151 Martin Boyer
9 INSURANCE FRAUD ESTIMATION: MORE EVIDENCE FROM THE QUEBEC
AUTOMOBILE INSURANCE INDUSTRY 175 Louis Caron and Georges Dionne
SECTION 3 Guest Speakers
10 THE SOCIETE DE L.:ASSURANCE AUTOMOBILE DU QUEBEC -
AN INTEGRATED MODEL OF ACTION TO INSURE AND PROTECT PEOPLE FROM RISKS INHERENT IN USE OF THE ROAD 183
Jean-Yves Gagnon
11 THEY CHEAT, YOU PAY! Raymond Medza
SECTION 4 Young and New Drivers: Licencing Policies, Evaluation and Risks
12 GRADUATED LICENSING IN QUEBEC: THE SEARCH FOR BALANCE
191
BETWEEN MOBILITY AND SAFETY 195 Claude Dussault and Patrice Letendre
viii AUTOMOBILE INSURANCE
5 COMMERCIAL VEHICLE INSURANCE: SHOULD FLEET POLICIES DIFFER
FROM SINGLE VEHICLE PLANS? 101 Claude Fluet
SECTION2 Insurance Fraud
6 COSTLY STATE FALSIFICATION OR VERIFICATION? THEORY AND
EVIDENCE FROM BODILY INJURY LlABILlTY CLAIMS Keith J. Crocker and Sharon Tennyson
7 THE FREQUENCY OF EXCESS CLAIMS FOR AUTOMOBILE PERSONAL
119
INJURIES 131 Alan F. Abrahamse and Stephen J. Carroll
8 WHEN IS THE PROPORTION OF CRIMINAL ELEMENTS IRRELEVANT?
A STUDY OF INSURANCE FRAUD WHEN INSURERS CANNOT COMMIT 151 Martin Boyer
9 INSURANCE FRAUD ESTIMATION: MORE EVIDENCE FROM THE QUEBEC
AUTOMOBILE INSURANCE INDUSTRY 175 Louis Caron and Georges Dionne
SECTION 3 Guest Speakers
10 THE SOCIETE DE L.:ASSURANCE AUTOMOBILE DU QUEBEC -
AN INTEGRATED MODEL OF ACTION TO INSURE AND PROTECT PEOPLE FROM RISKS INHERENT IN USE OF THE ROAD 183
Jean-Yves Gagnon
11 THEY CHEAT, YOU PAY! Raymond Medza
SECTION 4 Young and New Drivers: Licencing Policies, Evaluation and Risks
12 GRADUATED LlCENSING IN QUEBEC: THE SEARCH FOR BALANCE
191
BETWEEN MOBILlTY AND SAFETY 195 Claude Dussault and Patrice Letendre
Contents ix
13 AN EVALUATION OFTHE EFFECTS ON CRASHES OF THE 1991
LEGISLATIVE REFORM ON NEW LICENSEES IN QUEBEC 201 Urs Maag, Georges Dionne, Denise Desjardins, Stephane Messier and
Claire Laberge-Nadeau
14 LICENSING POLICIES FOR YOUNG DRIVERS IN THE UNITED STATES 215 Allan F. Williams
15 REDUCING THE RISK OF NEW DRIVERS THROUGH LEGISLATION AND
REGULATION 221 Dan Mayhew
16 EVALUATION OF THE ACCOMPANIED DRIVER TRAINING BY MEANS OF
A MARKOV CHAIN 231 Sylvain Lassarre and Pierre-Alain Hoyau
17 RISKY DRIVING BY YOUTH A. lames McKnight
18 YOUNG PEOPLE, ALCOHOL AND RISKS lean-Paul Assailly
SECTION 5 Road Insurance Regulation
19 NO FAULT AUTOMOBILE INSURANCE AND ACCIDENT SEVERITY:
LESSONS STILL TO BE LEARNED Rose Anne Devlin
20
243
253
267
THE INCENTIVE EFFECTS OF NO FAULT AUTOMOBILE INSURANCE 283 l. David Cummins and Mary A. Weiss
21 ESTIMATING THE EFFECTS OF "NO-PAY, NO-PLAY" AUTO INSURANCE
PLANS ON THE COSTS OF AUTO INSURANCE: THE EFFECTS OF PROPOSITION 213 309
Stephen l. Carroll and Allan F. Abrahamse
22 ANALYSIS OF THE ECONOMIC IMPACT OF MEDICAL AND OPTOMETRIC
DRIVING STANDARDS ON COSTS INCURRED BY TRUCKING FIRMS AND ON THE SOCIAL COSTS OF TRAFFIC ACCIDENTS 323
Georges Dionne. Claire Laberge-Nadeau. Denise Desjardins. Stephane Messier and Urs Maag
Contents ix
13 AN EVALUATION OFTHE EFFECTS ON CRASHES OFTHE 1991
LEGISLATIVE REFORM ON NEW LlCENSEES IN QUEBEC 201 Urs Maag, Georges Dionne, Denise Desjardins, Stephane Messier and
Claire Laberge-Nadeau
14 LlCENSING POLICIES FOR YOUNG DRIVERS IN THE UNITED STATES 215 Alian F. Williams
15 REDUCING THE RISK OF NEW DRIVERS THROUGH LEGISLATION AND
REGULATION 221 Dan Mayhew
16 EVALUATION OF THE ACCOMPANIED DRIVER TRAINING BY MEANS OF
A MARKOV CHAIN 231 Sylvain Lassarre and Pierre-Alain Hoyau
17 RISKY DRIVING BY YOUTH A. lames McKnight
18 YOUNG PEOPLE, ALCOHOL AND RISKS lean-Paul Assailly
SECTION 5 Road Insurance Regulation
19 NO FAULT AUTOMOBILE INSURANCE AND ACCIDENT SEVERITY:
LESSONS STILL TO BE LEARNED Rose Anne Dev/in
20
243
253
267
THE INCENTIVE EFFECTS OF NO FAULT AUTOMOBILE INSURANCE 283 1. David Cummins and Mary A. Weiss
21 ESTIMATING THE EFFECTS OF "NO-PAY, NO-PLAY" AUTO INSURANCE
PLANS ON THE COSTS OF AUTO INSURANCE: THE EFFECTS OF PROPOSITION 213 309
Stephen 1. Carroll and Alian F. Abrahamse
22 ANALYSIS OF THE ECONOMIC IMPACT OF MEDICAL AND OPTOMETRIC
DRIVING STANDARDS ON COSTS INCURRED BY TRUCKING FIRMS AND ON THE SOCIAL COSTS OF TRAFFIC ACCIDENTS 323
Georges Dionne. Claire Laberge-Nadeau. Denise Desjardins. Stephane Messier and Urs Maag
Guest Speakers and Contributing Authors
Guest Speakers
JEAN-YVES GAGNON President, Societe de l'Assurance Automobile du Quebec, Canada
Contributing Authors
ALLAN F. ABRAHAMSE Rand Corporation, USA allan_abrahamse@ rand. org
JEAN-PAUL ASSAILLY INRETS, France assailly@inrets.fr
DENIS BOLDUC Universite Laval, Canada dbol@ecn.ulaval.ca
SYLVIE BONIN Universite Laval, Canada sbon@grimes.ulaval.ca
MARTIN BOYER HEC-Montreal, Canada martin. boyer@hec.ca
LOUIS CARON Universite de Montreal, Canada caronlo@microtec.net
STEPHEN J. CARROLL Rand Corporation, USA sjc@rand.org
PIERRE-ANDRE CHIAPPORI University of Chicago, USA pchiappo@midway.uchicago.edu
KEITH J. CROCKER University of Michigan, USA kcrocker@umich.edu
J. DAVID CUMMINS University of Pennsylvania, USA cummins@ rider. wharton. upenn.edu
RAYMOND MEDZA General Manager, Insurance Bureau of Canada, Canada
DENISE DESJARDINS Universite de Montreal, Canada denise@crt.umontreal.ca
ROSE ANNE DEVLIN Universite d'Ottawa, Canada radevlin@uottawa.ca
GEORGES DIONNE HEC-Montreal, Canada dionne@crt. umontreal. ca
CLAUDE DUSSAULT Societe de I 'Assurance Automobile du Quebec, Canada claude.dussault@saaq.gouv.qc.ca
CLAUDE FLUET Universite du Quebec a Montreal, Canada fluet. claude-denys@uqam.ca
CHRISTIAN GOURIEROUX CREST-CEPREMAp, France gouriero@ensae·fr
PIERRE-ALAIN HOY AU INRETS, France hoyau@inrets.fr
CLAIRE LABERGE-NADEAU Universite de Montreal, Canada claire@crt.umontreal.ca
SYLVAIN LASSARRE INRETS, France lassarre@inrets.fr
Guest Speakers and Contributing Authors
Guest Speakers
JEAN-YVES GAGNON President, Societe de l'Assurance Automobile du Quebec, Canada
Contributing Authors
ALLAN F. ABRAHAMSE Rand Corporation, USA allan_abrahamse@ rand. org
JEAN-PAUL ASSAILLY INRETS, France assailly@inrets.Jr
DENIS BOLDUC Universite Laval, Canada dbol@ecn.ulaval.ca
SYLVIE BONIN Universite Laval, Canada sbon@grimes.ulaval.ca
MARTIN BOYER HEC-Montreal, Canada martin. boyer@hec.ca
LOUIS CARON Universite de Montreal, Canada caronlo@microtec.net
STEPHEN J. CARROLL Rand Corporation, USA sjc@rand.org
PIERRE-ANDRE CHIAPPORI University of Chicago, USA pchiappo@midway.uchicago.edu
KEITH J. CROCKER University of Michigan, USA kcrocker@umich.edu
J. DAVID CUMMINS University of Pennsylvania, USA cummins@ rider.wharton. upenn.edu
RAYMOND MEDZA General Manager, lnsurance Bureau of Canada, Canada
DENISE DESJARDINS Universite de Montreal, Canada denise@crt.umontreal.ca
ROSE ANNE DEVLIN Universite d'Ottawa, Canada radevlin@uottawa.ca
GEORGES DIONNE HEC-Montreal, Canada dionne@crt. umontreal. ca
CLAUDE DUSSAULT Sociite de 1 'Assurance Automobile du Quebec, Canada claude.dussault@saaq.gouv.qc.ca
CLAUDE FLUET Universite du Quebec il Montreal, Canada fluet. claude-denys@uqam.ca
CHRISTIAN GOURIEROUX CREST-CEPREMAp, France gouriero@ensae·fr
PIERRE-ALAIN HOYAU INRETS, France hoyau@inrets.Jr
CLAIRE LABERGE-NADEAU Universite de Montreal, Canada claire@crt.umontreal.ca
SYLVAIN LASSARRE INRETS, France lassarre@inrets.Jr
PATRICE LETENDRE Societe de I 'Assurance Automobile du Quebec, Canada patrice.letendre@saaq.gouv.qc.ca
URsMAAG Universite de Montreal, Canada maag@dms.umontreal.ca
DAN MAYHEW Traffic Injury Research Foundation, Canada danm@iosphere.net
A. JAMES McKNIGHT National Public Services Research Institute, USA jmcknigh@pirc.org
STEPHANE MESSIER Universite de Montreal, Canada stephane@crt.umontreal.ca
JEAN PINQUET Universite de Paris-X, France pinquet@u-parisIO·fr
SHARON TENNYSON Cornell University, USA st96@cornell.edu
CHARLES VANASSE Universite de Montreal, Canada vanasse@ibm.net
MARY A. WEISS Temple University, USA mweiss@vm.temple.edu
ALLAN F. WILLIAMS Insurance Institute for Highway Safety, USA awilliams@iihs.org
PATRICE LETENDRE Societe de 1 'Assurance Automobile du Quebec, Canada patrice.letendre@saaq.gouv.qc.ca
URsMAAG Universite de Montreal, Canada maag@dms.umontreal.ca
DANMAYHEW Traffic Injury Research Foundation, Canada danm@iosphere.net
A. JAMES McKNIGHT National Public Services Research Institute, USA jmcknigh@pirc.org
STEPHANE MESSIER Universite de Montreal, Canada stephane@crt.umontreal.ca
JEAN PINQUET Universite de Paris-X, France pinquet@u-parisIO·fr
SHARON TENNYSON Cornell University, USA st96@cornell.edu
CHARLES VANASSE Universite de Montreal, Canada vanasse@ibm.net
MARY A. WEISS Temple University, USA mweiss@vm.temple.edu
ALLAN F. WILLIAMS Insurance Institute for Highway Safety, USA awilliams@iihs.org
Referees
D. BOLDUC Universite Laval, Canada
L. CARON Universite de Montreal, Canada
M.CHIPMAN University of Toronto, Canada
K. CROCKER University of Michigan, USA
J. D. CUMMINS University of Pennsylvania, USA
M. DAGENAIS Universite de Montreal, Canada
R.A. DERRIG Automobile Insurers Bureau of Massachusetts, USA
R.A. DEVLIN Universite d'Ottawa, Canada
G.DIONNE HEC-Montreal, Canada
C. DUSSAULT Societe de I 'Assurance Automobile du Quebec, Canada
J.-P. FLORENS Universite des Sciences Sociales de Toulouse, France
C. FLUET Universite du Quebec Ii Montreal, Canada
S. HARRINGTON University of South Carolina. USA
B.JONAH Transport Canada, Canada
C. LABERGE-NADEAU Universite de Montreal, Canada
J. McKNIGHT National Public Services Research Institute, USA
S. MESSIER Universite de Montreal, Canada
J.MORGAN Princeton University, USA
y. PAGE Observatoire National Interministeriel de Securite Routiere, France
E. PETRUCELLI Association for the Advancement of Automotive Medicine, USA
P. PICARD Universite de Paris X Nanterre, France
F. PICHETTE Societe de {'Assurance Automobile du Quebec, Canada
J. PINQUET Universite de Paris X-Nanterre, France
B. PLESS Hopital de Montreal pour enfants, Canada
J. ROBERT Universiste de Montreal, Canada
B. SALANIE INSEE - CREST, France
H. SIMPSON Traffic Injury Research Foundation, Canada
A. SNOW The University of Georgia, USA
S. TENNYSON Cornell University, USA
P. WALLER UMTRI, USA
J. WILSON Insurance Corporation of British Columbia, Canada
R. WINTER University of Toronto, Canada
S. ZYWOKARTE Federal Highway Administration, USA
Referees
D. BOLDUC Universite Laval, Canada
L. CARON Universite de Montreal, Canada
M.CHIPMAN University of Toronto, Canada
K. CROCKER University of Michigan, USA
J. D. CUMMINS University of Pennsylvania, USA
M. DAGENAIS Universite de Montreal, Canada
R.A. DERRIG Automobile Insurers Bureau of Massachusetts, USA
R.A. DEVLIN Universite d'Ottawa, Canada
G.DIONNE HEC-Montreal, Canada
C. DUSSAULT Societe de l'Assurance Automobile du Quebec, Canada
J.-P. FLORENS Universite des Sciences Sociales de Toulouse, France
C. FLUET Universite du Quebec ii Montreal, Canada
S. HARRINGTON University of South Carolina. USA
B.JONAH Transport Canada, Canada
C. LABERGE-NADEAU Universite de Montreal, Canada
J. McKNIGHT National Public Services Research Institute, USA
S. MESSIER Universite de Montreal, Canada
J.MORGAN Princeton University, USA
Y. PAGE Observatoire National Interministeriel de Securite Routiere, France
E. PETRUCELLI Association for the Advancement of Automotive Medicine, USA
P. PICARD Universite de Paris X Nanterre, France
F. PICHETTE Societe de l'Assurance Automobile du Quebec, Canada
J. PINQUET Universite de Paris X-Nanterre, France
B. PLESS Hopital de Montreal pour enfants, Canada
J. ROBERT Universiste de Montreal, Canada
B. SALANIE INSEE - CREST, France
H. SIMPSON Traffic Injury Research Foundation, Canada
A. SNOW The University of Georgia, USA
S. TENNYSON Cornell University, USA
P. WALLER UMTRI, USA
J. WILSON Insurance Corporation of British Columbia, Canada
R. WINTER University of Toronto, Canada
s. ZYWOKARTE Federal Highway Administration, USA
Preface
TEODOR GABRIEL CRAINIC, DIRECfOR
The Center for Research on Transportation (C.R.T.) was founded in 1971 by the Universite de Montreal. From 1988 on, it is jointly managed by the Universite de Montreal and its affiliated schools, the Ecole des Hautes Etudes Commerciales and Ecole Poly technique. Professors, students and researchers from many institutions in the Montreal area join forces at the C.R.T. to analyze transportation, logistics and telecommunication systems from a multidisciplinary perspective.
The C.R.T. pursues three major, complimentary objectives: training of high-level specialists; the advancement of knowledge and technology; the transfer of technology towards industry and the public sector. Its main field of expertise is the development of quantitative and computer-based models and methods for the analysis of urban, regional and intercity transportation networks, as well as telecommunication systems. This applies to the study of passenger and commodity flows, as well as to the socioeconomic aspects of transportation: policies, regulation, economics.
The twenty-fifth anniversary of the C.R.T. offered the opportunity to evaluate past accomplishments and to identify future trends and challenges. Five colloquia were thus organized on major research and application themes that also reflected our main research areas. They gathered together internationally renowned researchers who linked recent scientific and technological advances to modeling and methodological challenges waiting to be tackled, particularly concerning new problems and applications, and the increasingly widespread use of new technologies.
The present book, together with its four companions, is the result of these meetings. I wish to thank my colleagues who organized these colloquia and also edited the books: PATRICE MARCaITE and SANG NGUYEN for Equilibrium and Advanced Transportation Modelling, BRUNILDE SANS6 and PATRICK SORIANO for Telecommunication Networks Planning, TEODOR GABRIEL CRAINIC and GILBERT LAPORTE for Fleet Management and Logistics, GEORGES DIONNE and CLAIRE LABERGE-NADEAU for Automobile Insurance: Road Safety, New Drivers, Risks, Insurance Fraud and Regulation and MARC GAUDRY and ROBERT MAYES for Taking Stock of Air Liberalization.
I also wish to thank all companies and institutions who financially supported the celebration of our twenty-fifth anniversary and the publication of the five books: BELL, BUREAU D' ASSURANCE DU CANADA, CANADIAN PACIFIC RAILWAY, ECOLE DES HAUTES ETUDES COMMERCIALES DE MONTREAL, INRO CONSULTANTS INC., LES ENTREPRISES GIRO INC., MINISTERE DES TRANSPORTS DU QuEBEC, SOCIETE DE L' ASSURANCE AUTOMOBILE DU QUEBEC, TRANSPORTS CANADA and the UNIVERSITE DE MONTREAL.
Preface
TEODOR GABRIEL CRAINIC, DIRECfOR
The Center for Research on Transportation (C.R.T.) was founded in 1971 by the Universite de Montreal. From 1988 on, it is jointly managed by the Universite de Montreal and its affiliated schools, the Ecole des Hautes Etudes Commerciales and Ecole Polytechnique. Professors, students and researchers from many institutions in the Montreal area join forces at the C.R.T. to analyze transportation, logistics and te1ecommunication systems from a multidisciplinary perspective.
The C.R.T. pursues three major, complimentary objectives: training of high-Ievel specialists; the advancement of knowledge and technology; the transfer of technology towards industry and the public sector. Its main field of expertise is the development of quantitative and computer-based models and methods for the analysis of urban, regional and intercity transportation networks, as well as telecommunication systems. This applies to the study of passenger and commodity flows, as well as to the socioeconomic aspects of transportation: policies, regulation, economics.
The twenty-fifth anniversary of the C.R.T. offered the opportunity to evaluate past accomplishments and to identify future trends and challenges. Five colloquia were thus organized on major research and application themes that also reflected our main research areas. They gathered together intemationally renowned researchers who linked recent scientific and technological advances to modeling and methodological challenges waiting to be tackled, particularly conceming new problems and applications, and the increasingly widespread use of new technologies.
The present book, together with its four companions, is the result of these meetings. 1 wish to thank my colleagues who organized these colIoquia and also edited the books: PATRICE MARCaITE and SANG NGUYEN for Equilibrium and Advanced Transportation Modelling, BRUNILDE SANS6 and PATRICK SORIANO for Telecommunication Networks Planning, TEODOR GABRIEL CRAINIC and GILBERT LAPORTE for F1eet Management and Logistics, GEORGES DIONNE and CLAIRE LABERGE-NADEAU for Automobile Insurance: Road Safety, New Drivers, Risks, Insurance Fraud and Regulation and MARc GAUDRY and ROBERT MAYES for Taking Stock of Air Liberalization.
1 also wish to thank alI companies and institutions who financially supported the celebration of our twenty-fifth anniversary and the publication of the five books: BELL, BUREAU D' ASSURANCE DU CANADA, CANADIAN PACIFIC RAILWAY, ECOLE DES HAUTES ETUDES COMMERCIALES DE MONTREAL, INRO CONSULTANTS INc., LES ENTREPRISES GIRO INc., MINISTERE DES TRANSPORTS DU QuEBEC, SOCIETE DE L' ASSURANCE AUTOMOBILE DU QUEBEC, TRANSPORTS CANADA and the UNIVERSITE DE MONTREAL.
Preface
TEODOR GABRIEL CRAINIC, DIRECTEUR
Le Centre de recherche sur les transports (C.R.T.) fut fonde en 1971 par I'Universite de Montreal. En 1988, deux institutions affiliees, l'Eco1e des Hautes Etudes Commerciales et I'Ecole Poly technique, se sont jointes a celle-ci pour former un centre multidisciplinaire conjoint. Des professeurs, etudiants et chercheurs provenant principalement des universites de la region montrealaise s'y regroupent pour mettre en commun leurs com¢tences diverses afin d'analyser les systemes de transport, logistiques et de telecommunication.
La mission du C.R.T. s'articule autour de trois axes complementaires: la formation de sp6cialistes de haut niveau; l'avancement des connaissances et des technologies; Ie transfert de ces technologies vers I' industrie et les organismes publics. L' expertise du c.R. T. est principalement associ6e au developpement de modeles et methodes quantitatifs et informatiques d'analyse des reseaux de transport urbains, regionaux, interurbains et internationaux ainsi que des reseaux de telecommunication. Celle-ci s'applique tout autant au transport de pas sagers et de marchandises qu'aux aspects socioeonomiques: reglementation, securite, economie du transport.
L'annee du vingt-cinquieme anniversaire nous a fourni l'occasion de faire Ie point et de nous tourner vers l'avenir. Cinq colloques portant sur des themes actuels et refletant les axes majeurs de recherche du C.R.T. sont issus de cette reflexion. Ces colloques, qui ont rassemble des chercheurs de reputation internationale, ont permis de discerner des liens entre les realisations recentes et les defis de modelisation et methodologiques qui nous attendent, particulierement dans les nouveaux champs de recherche et d'application, et dans l'utilisation grandissante de nouvelles technologies.
Ce livre et ses quatre compagnons sont Ie resultat tangible de ces colloques. Je remercie mes collegues qui les ont organises et animes et qui ont egalement produit ces livres : PATRICE MARCOTIE et SANG NGUYEN pour Equilibrium and Advanced Transportation Modelling, BRUNILDE SANS6 et PATRICK SORIANO pour Telecommunication Networks Planning, TEODOR GABRIEL CRAINIC et GILBERT LAPORTE pour Fleet Management and Logistics, GEORGES DIONNE et CLAIRE LABERGE-NADEAU pour Automobile Insurance : Road Safety, New Drivers, Risks, Insurance Fraud and Regulation et MARC GAUDRY et ROBERT MAYES pour Taking Stock of Air Liberalization.
Je tiens egalement a remercier Ies compagnies et institutions qui nous ont appuye financierement dans la realisation des celebrations du vingt-cinquieme anniversaire et la publication de ces livres : BELL, Ie BUREAU D' ASSURANCE DU CANADA, CANADIAN PACIFIC RAILWAY, l'ECOLE DES HAUTES ETUDES COMMERCIALES DE MONTREAL, LES CONSEILLERS INRO INC., LES ENTREPRISES GIRO INC., Ie MINISTERE DES TRANSPORTS DU QUEBEC, la SOCIETE DE L' ASSURANCE AUTOMOBILE DU QUEBEC, TRANSPORTS CANADA et l'UNIVERSITE DE MONTREAL.
Preface
TEODOR GABRIEL CRAINIC, DIRECTEUR
Le Centre de recherche sur les transports (C.R.T.) fut fonde en 1971 par I'Universite de Montreal. En 1988, deux institutions affiliees, l'Ecole des Hautes Etudes Commerciales et I'Ecole Polytechnique, se sont jointes a celle-ci pour former un centre multidisciplinaire conjoint. Des professeurs, etudiants et chercheurs provenant principalement des universites de la region montrealaise s'y regroupent pour mettre en commun leurs competences diverses afin d'analyser les systemes de transport, logistiques et de telecommunication.
La mission du C.R.T. s'articule autour de trois axes complementaires: la formation de sp6cialistes de haut niveau; l'avancement des connaissances et des technologies; le transfert de ces technologies vers l' industrie et les organismes publics. L' expertise du C.R. T. est principalement associ6e au developpement de modeles et methodes quantitatifs et informatiques d'analyse des reseaux de transport urbains, regionaux, interurbains et internationaux ainsi que des reseaux de telecommunication. Celle-ci s'applique tout autant au transport de passagers et de marchandises qu'aux aspects socioeonomiques: reglementation, securi te, economie du transport.
L'annee du vingt-cinquieme anniversaire nous a fourni l'occasion de faire le point et de nous tourner vers l'avenir. Cinq colloques portant sur des themes actuels et refletant les axes majeurs de recherche du C.R.T. sont issus de cette reflexion. Ces colloques, qui ont rassemble des chercheurs de reputation internationale, ont permis de discerner des liens entre les realisations recentes et les defis de modelisation et methodologiques qui nous attendent, particulierement dans les nouveaux champs de recherche et d'application, et dans l'utilisation grandissante de nouvelles technologies.
Ce livre et ses quatre compagnons sont le resultat tangible de ces colloques. Je remercie mes collegues qui les ont organises et animes et qui ont egalement produit ces livres : PATRICE MARCOTIE et SANG NGUYEN pour Equilibrium and Advanced Transportation ModeUing, BRUNILDE SANS6 et PATRICK SORIANO pour Telecommunication Networks Planning, TEODOR GABRIEL CRAINIC et GILBERT LAPORTE pour F1eet Management and Logistics, GEORGES DIONNE et CLAIRE LABERGE-NADEAU pour Automobile Insurance : Road Safety, New Drivers, Risks, Insurance Fraud and Regulation et MARC GAUDRY et ROBERT MAYES pour Taking Stock of Air Liberalization.
Je tiens egalement a remercier les compagnies et institutions qui nous ont appuye financierement dans la realisation des celebrations du vingt-cinquieme anniversaire et la publication de ces livres : BELL, le BUREAU D' ASSURANCE DU CANADA, CANADIAN PACIFIC RAILWAY, l'ECOLE DES HAUTES ETUDES COMMERCIALES DE MONTREAL, LES CONSEILLERS INRO INC., LES ENTREPRISES GIRO INC., le MINISTERE DES TRANSPORTS DU QUEBEC, la SOCIETE DE L' ASSURANCE AUTOMOBILE DU QUEBEC, TRANSPORTS CANADA et l'UNIVERSITE DE MONTREAL.
Foreword
J. DAVID CUMMINS, ADVISORY EDITOR
The automobile has become an indispensable means of transportation in the modern world. This is especially true in North America, where geography, population densities, and limited investments in public transportation make the automobile virtually the only local option for most people to get to work and conduct their lives. The heavy reliance on automobiles inevitably leads to large numbers of accidents, resulting in personal injuries and property damage. The costs of accidents are spread throughout the population of drivers through private (and in some instances public) insurance. The high costs of medical care and of repairing increasingly sophisticated vehicles has led to premium rates that represent a substantial share of disposable income in many parts of North America, and auto insurance prices have become a potent political issue in many states and Canadian provinces. These transportation realities motivated the University of Montreal's Center for Research on Transportation and the HEC Risk Management Chair to hold a conference on Automobile Insurance in April of 1997. The papers presented at the conference are published in this book.
Although the occurrence of automobile accidents is inevitable, the rate at which accidents occur and the costs of compensating accidents are not immutable but can be controlled through contracting, changes in the legal rules under which claims are settled, changes in driver training, licensing, and penalties for careless driving, and changes in the driving environment. This observation provides the overarching theme of chapters in this book - managing the transportation system to minimize the costs of accidents.
One set of chapters deals with the very important issue of contracting. The role of contracting is to control the problems of moral hazard and adverse selection that can cause insurance markets to fail and can reduce the level of economic welfare even if the market does not fail. Moral hazard and adverse selection result from information asymmetries between drivers/claimants and insurers about driver accident propensities ex ante and the existence and amount of damages resulting from accidents ex post. Several of the papers in this book deal with measuring or testing for adverse selection and moral hazard, and controlling or reducing information asymmetries through contract design and pricing strategies.
Although insurance is a valuable mechanism for improving economic welfare and sharing the costs of accidents, the existence of insurance itself raises the costs of accidents by weakening incentives for safe driving and providing incentives for filing false and inflated claims. The latter problem, insurance fraud, is exacerbated by poorly designed contracts and legal rules governing accident compensation and claims settlement. This book contains several excellent papers dealing with both theoretical and empirical aspects of insurance fraud. In addition, three papers deal with the effects of the accident
Foreword
J. DAVID CUMMINS, ADVISORY EDITOR
The automobile has become an indispensable means of transportation in the modern world. This is especially true in North America, where geography, population densities, and limited investments in public transportation make the automobile virtually the only local option for most people to get to work and conduct their lives. The heavy reliance on automobiles inevitably leads to large numbers of accidents, resulting in personal injuries and property damage. The costs of accidents are spread throughout the population of drivers through private (and in some instances public) insurance. The high costs of medical care and of repairing increasingly sophisticated vehicles has led to premium rates that represent a substantial share of disposable income in many parts of North America, and auto insurance prices have become a potent political issue in many states and Canadian provinces. These transportation realities motivated the University of Montreal's Center for Research on Transportation and the HEC Risk Management Chair to hold a conference on Automobile Insurance in April of 1997. The papers presented at the conference are published in this book.
Although the occurrence of automobile accidents is inevitable, the rate at which accidents occur and the costs of compensating accidents are not immutable but can be controlled through contracting, changes in the legal rules under which claims are settled, changes in driver training, licensing, and penalties for careless driving, and changes in the driving environment. This observation provides the overarching theme of chapters in this book - managing the transportation system to minimize the costs of accidents.
One set of chapters deals with the very important issue of contracting. The role of contracting is to control the problems of moral hazard and adverse selection that can cause insurance markets to fail and can reduce the level of economic welfare even if the market does not fail. Moral hazard and adverse selection result from inforrnation asymmetries between drivers/claimants and insurers about driver accident propensities ex ante and the existence and amount of damages resulting from accidents ex post. Several of the papers in this book deal with measuring or testing for adverse selection and moral hazard, and controlling or reducing inforrnation asymmetries through contract design and pricing strategies.
Although insurance is a valuable mechanism for improving economic welfare and sharing the costs of accidents, the existence of insurance itself raises the costs of accidents by weakening incentives for safe driving and providing incentives for filing false and inflated claims. The latter problem, insurance fraud, is exacerbated by poorlY designed contracts and legal rules governing accident compensation and claims settlement. This book contains several excellent papers dealing with both theoretical and empiric al aspects of insurance fraud. In addition, three papers deal with the effects of the accident
XX AUTOMOBILE INSURANCE
compensation system on insurance costs, discussing various aspects of the tort compensation system versus the principal alternative, no fault. A disproportionate share of accidents involves youthful drivers, and the book provides extensive discussions of policy changes that have been or could be made to reduce accidents among this segment of the popUlation.
This book is being published in two Kluwer book series: Centre for Research on Transportation, 25th Anniversary Series and the Huebner International Series on Risk, Insurance, and Economic Security. As Advisory Editor of the latter series, I am very pleased to be able to add this book to our list of offerings. It complements several of our other books which deal with insurance pricing, incentives, and contracting in another important area of insurance - workers compensation - and generally adds to the scope and depth of our book list. It is an outstanding contribution to the literature and the editors as well as the conference participants are to be commended for their hard work in producing so many high quality papers.
XX AUTOMOBILE INSURANCE
compensation system on insurance costs, discussing various aspects of the tort compensation system versus the principal alternative, no fault. A disproportionate share of accidents involves youthful drivers, and the book provides extensive discussions of policy changes that have been or could be made to reduce accidents among this segment of the population.
This book is being published in two Kluwer book series: Centre for Research on Transportation, 25th Anniversary Series and the Huebner International Series on Risk, Insurance, and Economic Security. As Advisory Editor of the latter series, I am very pleased to be able to add this book to our list of offerings. It complements several of our other books which deal with insurance pricing, incentives, and contracting in another important area of insurance - workers compensation - and generally adds to the scope and depth of our book list. It is an outstanding contribution to the literature and the editors as well as the conference participants are to be commended for their hard work in producing so many high quality papers.
INTRODUCTION Georges Dionne and Claire Laberge-Nadeau
Motor vehicle accidents are still a leading cause of death, even if the trend has somewhat declined over the past 20 years. The table in Appendix shows clearly that motor vehicle accidents are a significant cause of death in comparison with the air and space transport accidents, homicides and even HN infections, causes which are more often highlighted in the media. As we will see in this book, motor vehicle accidents are particularly damaging to very young drivers.
The assessment of driving risks is a common concern for road transportation safety and the automobile insurance industry. In both cases, there is an awareness of the great losses resulting from the death, injuries and property damage caused by traffic crashes. Research is essential to counteract this public health threat, to assess the success or failure of countermeasures, and to solve the problems it generates in the insurance industry.
The Centre for Research on Transportation (CRT) has been developing research activities on the above-mentioned issues since its inception. In 1988, these activities were strengthened and refined when, in collaboration with the Ministere des transports du Quebec, the CRT created its Laboratory on Transportation Safety. Since that date, the Laboratory's multidisciplinary group of professors, students and research professionals (epidemiologists, economists, statisticians, computer experts, engineers, psychologists, geographers, public health physicians) has been actively involved in several research projects - working in the subgroups of expertise each project requires. Most of the Laboratory's research is financed by the Societe de I'assurance automobile du Quebec (SAAQ), the Ministere des transports du Quebec (MTQ), with the collaboration of other funding agencies such as the Fonds pour laformation de chercheurs et I 'aide a la recherche (FeAR-Quebec), the Fonds de la recherche en sante du Quebec (FRSQ), the Social Sciences and Humanities Research Council of Canada (SSHRC) and also Transport Canada, the Insurance Bureau of Canada, the Jean Meloche Foundation, the Insurance Institute for Highway Safety, the American Automobile Association Foundation and the Federation Franraise des Societes d'Assurances (FFSA).
The contents of this book are in large part a reflection of many of the research and internationally collaborative activities carried out by the Laboratory over recent years. This is a book for people concerned about road crashes (prevention and compensation) and about the insurance problems they pose - namely private and public institutional authorities, consultants, administrators, practitioners, and researchers interested in sharing the authors' experience in this domain. The book presents original contributions related to motor vehicle insurance and road safety. All papers have been evaluated by external
INTRODUCTION Georges Dionne and Claire Laberge-Nadeau
Motor vehic1e accidents are still a leading cause of death, even if the trend has somewhat dec1ined over the past 20 years. The table in Appendix shows c1early that motor vehic1e accidents are a significant cause of death in comparison with the air and space transport accidents, homicides and even HN infections, causes which are more of ten highlighted in the media. As we will see in this book, motor vehic1e accidents are particularly damaging to very young drivers.
The assessment of driving risks is a common concern for road transportation safety and the automobile insurance industry. In both cases, there is an awareness of the great losses resulting from the death, injuries and property damage caused by traffic crashes. Research is essential to counteract this public health threat, to assess the success or failure of countermeasures, and to solve the problems it generates in the insurance industry.
The Centre for Research on Transportation (CRT) has been developing research activities on the above-mentioned issues since its inception. In 1988, these activities were strengthened and refined when, in collaboration with the Ministere des transports du Quebec, the CRT created its Laboratory on Transportation Safety. Since that date, the Laboratory's multidisciplinary group of professors, students and research professionals (epidemiologists, economists, statisticians, computer experts, engineers, psychologists, geographers, public health physicians) has been actively involved in several research projects - working in the subgroups of expertise each project requires. Most of the Laboratory's research is financed by the Societe de l'assurance automobile du Quebec (SAAQ), the Ministere des transports du Quebec (MTQ), with the collaboration of other funding agencies such as the Fonds pour la formation de chercheurs et l'aide il la recherche (FCAR-Quebec), the Fonds de la recherche en sante du Quebec (FRSQ), the Social Sciences and Humanities Research Council of Canada (SSHRC) and also Transport Canada, the Insurance Bureau of Canada, the Jean Meloche Foundation, the Insurance Institute for Highway Safety, the American Automobile Association Foundation and the Federation Franţaise des Societes d'Assurances (FFSA).
The contents of this book are in large part a reflection of many of the research and internationally collaborative activities carried out by the Laboratory over recent years. This is a book for people concerned about road crashes (prevention and compensation) and about the insurance problems they pose - namely private and public institutional authorities, consultants, administrators, practitioners, and researchers interested in sharing the authors' experience in this domain. The book presents original contributions related to motor vehic1e insurance and road safety. All papers have been evaluated by external
XXlJ AUTOMOBILE INSURANCE
referees and subsequently revised by the authors. Four subjects are covered: 1) Automobile Insurance Pricing, Risks, and Asymmetric Information; 2) Insurance Fraud; 3) Young Drivers: Licensing Policies, Evaluation and Risks; and 4) Road Insurance Regulation.
Automobile Insurance Pricing and Asymmetric Information
Automobile insurance started up as soon as automobiles first appeared on the roads. It became rapidly clear that this insurance needed special consideration due to the nature of road activity. More specifically, driving activities generate externalities and individual risks are difficult to evaluate and monitor since they are not directly observable.
Two information problems have been extensively discussed in economic and financial literatures since the beginning of the 1960's: 1) moral hazard referring to the effect insurance contracting has on the insured incentives whose actions are not perfectly observable by the insurer and, 2) adverse selection which refers to the effect private information about individual's type has on risks exchanges. These two problems have been used intensively as stylized facts to explain the existence of partial insurance coverage (deductible and coinsurance), risk classification, and bonus-malus. However, very few empirical studies have been able to verify adequately the real significance of the two problems. This lack of results is explained by two factors: the non-availability of micro data and the difficulty of setting up the adequate methodologies.
The first paper, by Chiappori, is about the methodological difficulties of isolating the significance of the two information problems in insurance markets. The objective is to review the empirical models published in the recent literature. Some of the main theoretical issues are also discussed. One important conclusion of this chapter is that multiperiod data seem necessary to separate the effects of moral hazard from those of adverse selection. Another conclusion is that the results can be affected by the methods used. In particular, some wrong conclusions on the presence of adverse selection may be due to specification problems in the econometric modeling.
Dionne-Gourieroux-Vanasse propose an empirical analysis of the presence of adverse selection in an insurance market. They first present a basic model of a market with adverse selection and then extend the theoretical model by introducing different issues related to transaction costs, accident costs, risk aversion and moral hazard. They propose specification tests that may be useful in isolating the presence of both residual adverse selection and residual moral hazard in the portfolio of an insurer. They apply their model to the data of a private insurer and they show that there is no residual adverse selection in the portfolio studied, since appropriate risk classification is made by the insurer. Consequently, the insurer does not need a self-selection mechanism, as the deductible choice, to reduce the effect of adverse selection.
The chapter by Jean Pinquet is also related to information problems. The author presents statistical models that permit experience rating in insurance. The chapter proposes consistent estimators for models related to both the number and the cost of claims. Examples are given for count data models. Empirical results are obtained from a French data base of automobile insurance contracts.
One difficult exercise in the prevention of road accidents is to detect dangerous road sites. The most widely used detection tools have a Bayes background. The chapter by
XXlJ AUTOMOBILE INSURANCE
referees and subsequently revised by the authors. Four subjects are covered: 1) Automobile Insurance Pricing, Risks, and Asymmetric Information; 2) Insurance Fraud; 3) Young Drivers: Licensing Policies, Evaluation and Risks; and 4) Road Insurance Regulation.
Automobile Insurance Pricing and Asymmetric Information
Automobile insurance started up as soon as automobiles first appeared on the roads. It became rapidly c1ear that this insurance needed special consideration due to the nature of road activity. More specifically, driving activities generate extemalities and individual risks are difficult to evaluate and monitor since they are not directly observable.
Two information problems have been extensively discussed in economic and financial literatures since the beginning of the 1960's: 1) moral hazard referring to the effect insurance contracting has on the insured incentives whose actions are not perfectly observable by the insurer and, 2) adverse selection which refers to the effect private information about individual's type has on risks exchanges. These two problems have been used intensively as stylized facts to explain the existence of partial insurance coverage (deductible and coinsurance), risk c1assification, and bonus-malus. However, very few empirical studies have been able to verify adequately the real significance of the two problems. This lack of results is explained by two factors: the non-availability of micro data and the difficulty of setting up the adequate methodologies.
The first paper, by Chiappori, is about the methodological difficulties of isolating the significance of the two information problems in insurance markets. The objective is to review the empirical models published in the recent literature. Some of the main theoretical issues are also discussed. One important conc1usion of this chapter is that multiperiod data seem necessary to separate the effects of moral hazard from those of adverse selection. Another conc1usion is that the results can be affected by the methods used. In particular, some wrong conc1usions on the presence of adverse selection may be due to specification problems in the econometric mode ling.
Dionne-Gourieroux-Vanasse propose an empirical analysis of the presence of adverse selection in an insurance market. They first present a basic model of a market with adverse selection and then extend the theoretical model by introducing different issues related to transaction costs, accident costs, risk aversion and moral hazard. They propose specification tests that may be useful in isolating the presence of both residual adverse selection and residual moral hazard in the portfolio of an insurer. They apply their model to the data of a private insurer and they show that there is no residual adverse selection in the portfolio studied, since appropriate risk c1assification is made by the insurer. Consequently, the insurer does not need a self-selection mechanism, as the deductible choice, to reduce the effect of adverse selection.
The chapter by Jean Pinquet is also related to information problems. The author presents statistical models that permit experience rating in insurance. The chapter proposes consistent estimators for models related to both the number and the cost of c1aims. Examples are given for count data models. Empirical results are obtained from a French data base of automobile insurance contracts.
One difficult exercise in the prevention of road accidents is to detect dangerous road sites. The most widely used detection tools have a Bayes background. The chapter by
Introduction XXlll
Bolduc and Bonin extends the general empirical Bayes models to the multinomial case. Their approach is a full-information Bayes method that allows for both deterministic and random heterogeneity as well as spatial-correlation among the sites under investigation. An empirical example based on Quebec accident data is provided.
Up to now, the design of insurance contracting for road accidents has been limited to individual vehicles. Fleet policies were not widely developed. The study by Fluet analyses the effect of fleet size on the design of insurance contracts under adverse selection and moral hazard. The author shows that if insurers have perfect ex-post information with respect to the occurrence of losses, the contract is almost fIrst-best if the fleet is sufficiently large. However, if the insurers have to rely on the insured for the reporting of losses, the contract is characterized by a ceiling on coverage and the fIrst-best approximation result cannot be guaranteed. For reasonably large fleets, the efficiency loss due to asymmetric information may be small.
Insurance Fraud
Insurance fraud is now a signifIcant resource allocation problem in many insurance markets. Its magnitUde is often documented as representing more than 10% of the value of automobile insurance claims. Many factors can explain this phenomenon, but its principal cause is due to information problems between the insureds and the insurers.
Crocker and Tennyson study the nature of ex-post moral hazard associated with the magnitude of losses. They consider two alternative explanations. Costly state verifIcation is a situation where only the insured knows the true magnitude of loss and the insurer must pay a fIxed monitoring cost to have access to the information. Usually, this type of information problem explains the presence of an insurance contract that minimizes the audit costs while compensating for large losses. Under costly verifIcation, the insured is able to modify an observed claim at a cost. The nature of an optimal contract in this case is less standard: it must balance the need for insurance coverage and the incentives that reduce falsifIcation.
Crocker and Tennyson investigate in detail these two information problems and derive the corresponding optimal insurance contracts. For costly verifIcation the optimal contract involves auditing and full payment of all claims that exceed a threshold level. For costly falsifIcation, the optimal contract involves the overpayment of claims for amounts below some specifIc threshold value, and underpayment of claims for those above the threshold. They also test their models by using US data on bodily liability settlements in automobile insurance.
Abrahamse and Carroll also study the problem of excess claims for automobile personal injuries. Their objective is to estimate the extend of excess claiming in the United States. In their study, excess claiming means "a claim for an alleged injury that is either nonexistent or unrelated to the accident". This defInition includes planned fraud and opportunistic fraud. They do not consider the problem of buildup. Their results indicate that different insurance systems modify the incentive to make excess claims for auto injuries. No-fault systems reduce these incentives; particularly, the verbal threshold no-fault systems appear to eliminate them. Finally, they estimated that about 28 percent of all claims submitted by auto accident victims are exaggerated.
Introduction XXlll
Bolduc and Bonin extends the general empirical Bayes models to the multinomial case. Their approach is a full-information Bayes method that allows for both deterministic and random heterogeneity as weU as spatial-correlation among the sites under investigation. An empiric al example based on Quebec accident data is provided.
Up to now, the design of insurance contracting for road accidents has been limited to individual vehicles. Fleet policies were not widely deve1oped. The study by Fluet analyses the effect of fleet size on the design of insurance contracts under adverse selection and moral hazard. The author shows that if insurers have perfect ex-post information with respect to the occurrence of losses, the contract is almost frrst-best if the fleet is sufficiently large. However, if the insurers have to rely on the insured for the reporting of losses, the contract is characterized by a ceiIing on coverage and the first-best approximation result cannot be guaranteed. For reasonably large fleets, the efficiency loss due to asymmetric information may be small.
Insurance Fraud
Insurance fraud is now a significant resource allocation problem in many insurance markets. Its magnitude is of ten documented as representing more than 10% of the value of automobile insurance claims. Many factors can explain this phenomenon, but its principal cause is due to information problems between the insureds and the insurers.
Crocker and Tennyson study the nature of ex-post moral hazard associated with the magnitude of losses. They consider two alternative explanations. Costly state verification is a situation where only the insured knows the true magnitude of loss and the insurer must pay a fixed monitoring cost to have access to the information. Usually, this type of information problem explains the presence of an insurance contract that minimizes the audit costs while compensating for large losses. Under costly verification, the insured is able to modify an observed claim at a cost. The nature of an optimal contract in this case is less standard: it must balance the need for insurance coverage and the incentives that reduce falsification.
Crocker and Tennyson investigate in detail these two information problems and derive the corresponding optimal insurance contracts. For costly verification the optimal contract involves auditing and full payment of aU claims that exceed a threshold level. For costly falsification, the optimal contract involves the overpayment of claims for amounts below some specific threshold value, and underpayment of claims for those above the threshold. They also test their models by using US data on bodily liability settlements in automobile insurance.
Abrahamse and Carroll also study the problem of excess claims for automobile personal injuries. Their objective is to estimate the extend of excess claiming in the United States. In their study, excess claiming means "a claim for an alleged injury that is either nonexistent or unrelated to the accident". This definition includes planned fraud and opportunistic fraud. They do not consider the problem of buildup. Their results indicate that different insurance systems modify the incentive to make excess claims for auto injuries. No-fault systems reduce these incentives; particularly, the verbal threshold no-fault systems appear to eliminate them. Finally, they estimated that about 28 percent of aU claims submitted by auto accident victims are exaggerated.
xxiv AUTOMOBILE INSURANCE
Caron and Dionne are involved in an estimation of insurance fraud but with a different method. Their problem is to estimate the total fraud in a given market when the available data are limited to fraud observed by the claim adjusters. Their data was limited to property damages. They show that a multiplicative factor of 3.4% is present in the data, which means that total fraud payments range from 96.2 to 208.4 million dollars while the observed fraud range from 28 to 61 million dollars in the Quebec automobile insurance industry (1994). Their Best Guess Estimator yields a 10% fraud rate in the total cost of claims. An interesting corollary is that the claim adjusters who participated in the survey (representing 70% of the market) observed only 1/3 of the potential frauds in the closed files studied. The authors interpret this number as an index of efficiency for the entire verification process in the industry. A natural question is: Why is this index of effi-ciency so low? .
One of our guest speakers, Mr. R. Medza, general manager of the Insurance Bureau of Canada, Section Quebec, does not answer to this question, but indicates that insurance fraud costs 2.3 billion dollars annually to the insurance consumers in Canada. The Canadian industry created a National Task Force to address insurance fraud. More recently, in 1994, a broader working group, the Canadian Coalition Against Insurance Fraud, was set up. Its five areas of activities concern: insurance delivery, investigation, laws and regulations, measurement and research, and public awareness. A particular attention is allowed to the public awareness regarding fraud.
Martin Boyer addresses a difficult problem related to insurance contracting in a context of potential fraud. Indeed, the significance of fraud may be explained by the noncommitment of insurers to their initial insurance contracts. For example, when the observed contract is a straight deductible, the insurer is committed to audit all claims above the deductible. It is clear that they do not audit all these claims, since audit is costly. Consequently, when this behavior is anticipated by the insureds, they have less incentive to tell the truth when they file a claim.
Boyer shows that a separating equilibrium cannot exist. Moreover, if the proportion of Honests is small enough in the insurer's portfolio, then the optimal contract is exactly the same as the one the Criminals will buy in a full information situation. A corollary of this result is that both the amount of fraud and the amount of detected fraud are independent of the exact proportion of Criminals provided that there are enough Criminals in the economy.
Young and New Drivers: Licensing Policies, Evaluation and Risks
The following seven articles are focussed on young and new drivers. This group of drivers has a relatively high rate of accidents. Indeed, in Quebec, in 1992, young drivers 16 to 24 years old were involved in 24% of the accidents resulting of bodily injuries, while they represented 13% oflicense holders, as documented by Dussault and Letendre. The results from the Maag et al. study confirm this tendency but show that differences between ages are a significant factor for accidents even during the first year of driving. This implies that new drivers do not represent an homogeneous group. Their common absence of experience is a significant factor of accidents for all ages of novice drivers, but the younger drivers register much more accidents, specifically the male young drivers of 16-17 years old. Other results related to the Quebec 1991 reform on licensing are presented. Note that this reform has been substantially modified in 1997.
xxiv AUTOMOBILE INSURANCE
Caron and Dionne are involved in an estimation of insurance fraud but with a different method. Their problem is to estimate the total fraud in a given market when the available data are limited to fraud observed by the claim adjusters. Their data was limited to property damages. They show that a multiplicative factor of 3.4% is present in the data, which means that total fraud payments range from 96.2 to 208.4 million dollars while the observed fraud range from 28 to 61 million dollars in the Quebec automobile insurance industry (1994). Their Best Guess Estimator yields a 10% fraud rate in the total cost of claims. An interesting corolIary is that the claim adjusters who participated in the survey (representing 70% of the market) observed only 1/3 of the potential frauds in the closed files studied. The authors interpret this number as an index of efficiency for the entire verification process in the industry. A natural question is: Why is this index of effi-ciency so low? .
One of our guest speakers, Mr. R. Medza, general manager of the Insurance Bureau of Canada, Section Quebec, does not answer to this question, but indicates that insurance fraud costs 2.3 billion dollars annualIy to the insurance consumers in Canada. The Canadian industry created a National Task Force to address insurance fraud. More recently, in 1994, a broader working group, the Canadian Coalition Against Insurance Fraud, was set up. Its five areas of activities concern: insurance delivery, investigation, laws and regulations, measurement and research, and public awareness. A particular attention is alIowed to the public awareness regarding fraud.
Martin Boyer addresses a difficult problem related to insurance contracting in a context of potential fraud. Indeed, the significance of fraud may be explained by the noncommitment of insurers to their initial insurance contracts. For example, when the observed contract is a straight deductible, the insurer is committed to audit alI claims above the deductible. It is clear that they do not audit alI these claims, since audit is costly. Consequently, when this behavior is anticipated by the insureds, they have less incentive to tell the truth when they file a claim.
Boyer shows that a separating equilibrium cannot exist. Moreover, if the proportion of Honests is small enough in the insurer's portfolio, then the optimal contract is exactly the same as the one the Criminals will buy in a full information situation. A corollary of this result is that both the amount of fraud and the amount of detected fraud are independent of the exact proportion of Criminals provided that there are enough Criminals in the economy.
Young and New Drivers: Licensing Policies, Evaluation and Risks
The folIowing seven articles are focussed on young and new drivers. This group of drivers has a relatively high rate of accidents. Indeed, in Quebec, in 1992, young drivers 16 to 24 years old were involved in 24% of the accidents resulting of bodily injuries, while they represented 13% oflicense holders, as documented by Dussault and Letendre. The results from the Maag et al. study confirm this tendency but show that differences between ages are a significant factor for accidents even during the first year of driving. This implies that new drivers do not represent an homogeneous group. Their common absence of experience is a significant factor of accidents for all ages of novice drivers, but the younger drivers register much more accidents, specificalIy the male young drivers of 16-17 years old. Other results related to the Quebec 1991 reform on licensing are presented. Note that this reform has been substantialIy modified in 1997.
Introduction xxv
The 1991 reform was designed to give new licensees more experience and better training before licensing. One objective of the study by Maag et al. was to evaluate the effect of this reform on crash rates during the first year after licensing. They found that the 1991 reform had no significant effect on crash rates. They also found that licensees who passed the three parts of the theory exam at the first attempt had significantly lower crash rates than those who needed more than one attempt to pass. Experience, in terms of the number of periods since obtaining the license, also had a significant effect. Average rates of crashes for the first three months for women and the first four months for men were higher than the rates for the subsequent months.
Dussault and Letendre discuss the effectiveness of the different countermeasures, available to prevent accidents, while analyzing their impact on the mobility of young drivers. Of the eleven measures examined, three became part of the new graduated licensing reform in Quebec (1997), precisely because they foster a better balance between mobility and safety. They are: 1) Zero alcohol; 2) Ceiling of 4 demerit points; and 3) Learner's license for 12 months.
Allan William's article shows that the trends observed in Quebec are similar to that in American states. His main objective is to analyze the role of graduated licensing where new drivers are encouraged to drive in lower-risk situations before full privileges are given. The introduction of graduated licensing seems to be efficient in reducing accidents, but it does introduce mobility costs. So, the regulator must find a fair trade-off between mobility and safety, a conclusion that agrees with the framework used for the set-up of new regulations in Quebec, as documented by Dussault and Letendre.
Dan Mayhew discusses different forms of regulation in force in several countries and their effects. Several jurisdictions have introduced graduated licensing and many others are considering doing so. The review of graduated licensing programs reveals that each of these programs is unique. However, there are also important similarities across programs. Experience in some jurisdictions suggests that a graduated licensing program may result in at least a 6 to 8 % reduction in collisions.
James McKnight's article analyses the extreme high crash risk of young drivers as a result of both inexperience and immaturity. The primary source of the problem is failure to perceive the degree of risk presented by the driving environment and to respond to it. Speed, distractions and use of alcohol are three subjects among those discussed by the author.
The next two articles are related to experiences conducted in France. Lassarre and Royau evaluated the accompanied driver training program by using a Markov chain. Two thousand drivers between 20 and 22 were questioned on: 1) the type of driver training they had undertaken in order to obtain a driving license; 2) their accident record over a three year period; and 3) other control variables. They obtained that the conditional probability of accident appears to increase with age and experience. This result contradicts the reduction in risk which has been observed in other studies. However, they did not have data on very young drivers. They also obtain that the training program seems to have a significative effect on accident risks. In their conclusion, the authors emphasize the fact that the quality of their sample is not fully reliable.
Introduction xxv
The 1991 reform was designed to give new licensees more experience and better training before licensing. One objective of the study by Maag et al. was to evaluate the effect of this reform on crash rates during the first year after licensing. They found that the 1991 reform had no significant effect on crash rates. They also found that licensees who passed the three parts of the theory exam at the first attempt had significantly lower crash rates than those who needed more than one attempt to pass. Experience, in terms of the number of periods since obtaining the license, also had a significant effect. Average rates of crashes for the first three months for women and the first four months for men were higher than the rates for the subsequent months.
Dussault and Letendre discuss the effectiveness of the different countermeasures, available to prevent accidents, while analyzing their impact on the mobility of young drivers. Of the eleven measures examined, three became part of the new graduated licensing reform in Quebec (1997), precisely because they foster a better balance between mobility and safety. They are: 1) Zero alcohol; 2) Ceiling of 4 demerit points; and 3) Leamer's license for 12 months.
Allan William's article shows that the trends observed in Quebec are similar to that in American states. His main objective is to analyze the role of graduated licensing where new drivers are encouraged to drive in lower-risk situations before full privileges are given. The introduction of graduated licensing seems to be efficient in reducing accidents, but it does introduce mobility costs. So, the regulator must find a fair trade-off between mobility and safety, a conclusion that agrees with the framework used for the set-up of new regulations in Quebec, as documented by Dussault and Letendre.
Dan Mayhew discusses different forms of regulation in force in several countries and their effects. Several jurisdictions have introduced graduated licensing and many others are considering doing so. The review of graduated licensing programs reveals that each of these programs is unique. However, there are also important similarities across programs. Experience in some jurisdictions suggests that a graduated licensing program may result in at least a 6 to 8 % reduction in collisions.
James McKnight's article analyses the extreme high crash risk of young drivers as a result of both inexperience and immaturity. The primary source of the problem is failure to perceive the degree of risk presented by the driving environment and to respond to it. Speed, distractions and use of alcohol are three subjects among those discussed by the author.
The next two articles are related to experiences conducted in France. Lassarre and Hoyau evaluated the accompanied driver training program by using a Markov chain. Two thousand drivers between 20 and 22 were questioned on: 1) the type of driver training they had undertaken in order to obtain a driving license; 2) their accident record over a three year period; and 3) other control variables. They obtained that the conditional probability of accident appears to increase with age and experience. This result contradicts the reduction in risk which has been observed in other studies. However, they did not have data on very young drivers. They also obtain that the training program seems to have a significative effect on accident risks. In their conclusion, the authors emphasize the fact that the quality of their sample is not fully reliable.
xxvi AUTOMOBILE INSURANCE
J.P. Assailly studied the decision to drive under the influence of alcohol in young drivers. Although it is well known that the alcohol-factor is a source of accident risk for the young, it is not clear that this population is homogeneous: not all young people spend their evenings in bars; not all young people drink the same quantity; and for the same blood alcohol concentration, not all young people react in the same way with regard to driving activities.
Some aspects of drinking and driving have to do with risk taking. For some young people, alcohol is something which is taken intentionally as an intoxicant. The preventive strategies should consider three different aspects of risk: some might stress the risk taking; others might focus on the perceived risk; and, finally, others might target the accepted risk. Prevention should take into account the complexity of alcohol's effects on driving behavior.
Road Insurance Regulation
The effect of No-Fault automobile insurance is still a matter of debate in many countries. Very few insurance regimes are pure no-fault. The Quebec public regime for bodily injuries is one of them. Now in place for 20 years, it is managed by a public monopoly insurer, the Societe de l' Assurance Automobile du Quebec (SAAQ). Our second guest speaker, Mr. J.Y. Gagnon, presents the SAAQ. For him, the SAAQ is not just a traditional insurer but more an "integrated model of action to insure and protect people from risks inherent in use of the road". In fact, the SAAQ is in charge of road safety in Quebec and the results are impressive. From 1980 to 1992, the province of Quebec had one of the most significant declines in fatalities in all industrialized countries. The financial results are also impressive. Over the last 20 years, the cost of premiums for bodily injuries has decreased for more than 50 percent, when inflation is taken into account. Finally, a recent study shows that compensation levels are just as generous as what could be obtained under a liability system.
R.A. Devlin studies the effects of no-fault insurance rules on the severity of accidents. Her analysis adds further ingredients to the broader issue concerning the incentive effects of no-fault automobile regimes, a subject that is also studied by Cummins and Weiss. The main result of Devlin's study is that no-fault rules in the United States do matter when it comes to the severity of injuries: the probability of sustaining a more serious accident is higher in a no-fault state than in a liability-only state. For the author, this means that drivers appear to take less care in no-fault states in comparison to liabilityonly ones.
Cummins and Weiss obtain that no-fault is associated with higher fatal accident rates than tort, with the strongest effect in states using verbal thresholds. Their data is for the 1982-1993 period. Consequently, for the authors, there is a trade-off between cost control and fatality rates when evaluating no fault proposals. However, their results also suggest that other mechanisms such as experience rating may provide effective alternatives to the tort system as incentives for safer driving. The SAAQ did introduce, in 1992, a new bonus-malus scheme in Quebec based on demerit points. Preliminary results show that this mechanism reduced both the number of accidents and the number of traffic violations as documented by J.M. Gagnon.
xxvi AUTOMOBILE INSURANCE
J.P. AssailIy studied the decision to drive under the influence of alcohol in young drivers. Although it is well known that the alcohol-factor is a source of accident risk for the young, it is not clear that this population is homogeneous: not alI young people spend their evenings in bars; not alI young people drink the same quantity; and for the same blood alcohol concentration, not alI young people react in the same way with regard to driving activities.
Some aspects of drinking and driving have to do with risk taking. For some young people, alcohol is something which is taken intentionalIy as an intoxicant. The preventive strategies should consider three different aspects of risk: some might stress the risk taking; others might focus on the perceived risk; and, finalIy, others might target the accepted risk. Prevention should take into account the complexity of alcohol's effects on driving behavior.
Road Insurance Regulation
The effect of No-Fault automobile insurance is still a matter of debate in many countries. Very few insurance regimes are pure no-fault. The Quebec public regime for bodily injuries is one of them. Now in place for 20 years, it is managed by a public monopoly insurer, the Societe de l' Assurance Automobile du Quebec (SAAQ). Our second guest speaker, Mr. J.Y. Gagnon, presents the SAAQ. For him, the SAAQ is not just a traditional insurer but more an "integrated model of action to insure and protect people from risks inherent in use of the road". In fact, the SAAQ is in charge of road safety in Quebec and the results are impressive. From 1980 to 1992, the province of Quebec had one of the most significant declines in fatalities in alI industrialized countries. The financial results are also impressive. Over the Iast 20 years, the cost of premiums for bodily injuries has decreased for more than 50 percent, when inflation is taken into account. FinalIy, a recent study shows that compensation levels are just as generous as what could be obtained under a liability system.
R.A. Devlin studies the effects of no-fault insurance rules on the severity of accidents. Her analysis adds further ingredients to the broader issue concerning the incentive effects of no-fault automobile regimes, a subject that is also studied by Cummins and Weiss. The main result of Devlin's study is that no-fault rules in the United States do matter when it comes to the severity of injuries: the probability of sustaining a more serious accident is higher in a no-fault state than in a liability-only state. For the author, this means that drivers appear to take less care in no-fault states in comparison to liabilityonly ones.
Cummins and Weiss obtain that no-fault is associated with higher fatal accident rates than tort, with the strongest effect in states using verbal thresholds. Their data is for the 1982-1993 period. Consequently, for the authors, there is a trade-off between cost control and fatality rates when evaluating no fault proposaIs. However, their results also suggest that other mechanisms such as experience rating may provide effective alternatives to the tort system as incentives for safer driving. The SAAQ did introduce, in 1992, a new bonus-malus scheme in Quebec based on demerit points. Preliminary results show that this mechanism reduced both the number of accidents and the number of traffic violations as documented by J.M. Gagnon.
Introduction xxvii
California voters adopted Proposition 213 in 1996. This new approach intends to cut the cost of automobile insurance. This legislation, identified as "no pay, no play", would limit uninsured motorists' rights to recovery of losses resulting from an automobile accident. Proposition 213 differs from previous strategies for reducing accident costs such as no-fault, by limiting the compensation rights of people who "were breaking the law when they were injured".
Now the question is: Will Proposition 213 reduce insurance premiums? In their chapter, Carroll and Abrahamse estimate what effects the Proposition's provisions on uninsured or drunk drivers will likely have on the costs of private passenger auto insurance. They obtain that a limited "no pay, no play" plan could reduce auto insurance costs up to 10 to 12 percent if current claiming, negotiating and insurance patterns persist; or it could reduce the auto insurance premiums by about 5 percent.
Truck accidents generate strong externalities. For each truck driver killed, an average of at least other six individuals are killed outside the truck. Consequently, the regulation of truck driving is an important issue. One such regulation concerns the medical conditions of truck drivers. Recent studies do not agree on the possible relationships between medical conditions and traffic safety. Moreover, most of them do not control for exposure factors. In their study, Dionne et al. obtain that diabetics truck drivers not in class 1, articulated trucks (79% of them are in class 3, straight trucks only) have more accidents than drivers in good health. No other medical condition studied has a significant effect on individual accident rates. Many risk exposure factors are significant. Dionne et al. also studied the severity of accidents in terms of the number of victims injured or killed. The results indicate that drivers with a visual impairment have more serious accidents than those in good health. Their cost estimations show that the expected average costs (private and social) of drivers with diabetes is twice as high as the expected average costs of drivers in good health. Differences are less significant for the drivers with a visual impairment.
Acknowledgements
We wish to express our gratitude to HEC-Montreal and the Huebner Foundation for Insurance Education at the University of Pennsylvania for providing financial support to the publication of this book. Thanks are also addressed to all the authors and referees for their significant contribution. It would have been difficult to produce this book without the generosity of our many collaborators at the Risk Management Chair at HEC and the Centre for Research on Transportation at the Universite de Montreal, who spent many hours organizing the conference and the publication process. We want to draw special attention to the collaboration of Claire Boisvert, Lucie Cournoyer, Gilles Gagnon, Marie-Gloriose Ingabire, Lucie L'Heureux, Remi Moreau, Clairette Simard and Claudine St-Pierre. Finally, much of the book's production was carried out in the graphics department at HEC. We thank Mireille Donais for her remarkable collaboration on this project.
Montreal, September 1998
lntroduction xxvii
California voters adopted Proposition 213 in 1996. This new approach intends to cut the cost of automobile insurance. This legislation, identified as "no pay, no play", would limit uninsured motorists' rights to recovery of losses resulting from an automobile accident. Proposition 213 differs from previous strategies for reducing accident costs such as no-fault, by limiting the compensation rights of people who "were breaking the law when they were injured".
Now the question is: Will Proposition 213 reduce insurance premiums? In their chapter, Carroll and Abrahamse estimate what effects the Proposition's provisions on uninsured or drunk drivers willlikely have on the costs of private passenger auto insurance. They obtain that a limited "no pay, no play" plan could reduce auto insurance costs up to 10 to 12 percent if current claiming, negotiating and insurance patterns persist; or it could reduce the auto insurance premiums by about 5 percent.
Truck accidents generate strong externalities. For each truck driver killed, an average of at least other six individuals are killed outside the truck. Consequently, the regulation of truck driving is an important issue. One such regulation concerns the medical conditions of truck drivers. Recent studies do not agree on the possible relationships between medical conditions and traffic safety. Moreover, most of them do not control for exposure factors. In their study, Dionne et al. obtain that diabetics truck drivers not in class 1, articulated trucks (79% of them are in class 3, straight trucks only) have more accidents than drivers in good health. No other medical condition studied has a significant effect on individual accident rates. Many risk exposure factors are significant. Dionne et al. also studied the severity of accidents in terms of the number of victims injured or killed. The results indicate that drivers with a vis ual impairment have more serious accidents than those in good health. Their cost estimations show that the expected average costs (private and social) of drivers with diabetes is twice as high as the expected average costs of drivers in good health. Differences are less significant for the drivers with a vis ual impairment.
Acknowledgements
We wish to express our gratitude to HEC-Montreal and the Huebner Foundation for Insurance Education at the University of Pennsylvania for providing financial support to the publication of this book. Thanks are also addressed to all the authors and referees for their significant contribution. It would have been difficult to produce this book without the generosity of our many collaborators at the Risk Management Chair at HEC and the Centre for Research on Transportation at the Universite de Montreal, who spent many hours organizing the conference and the publication process. We want to draw special attention to the collaboration of Claire Boisvert, Lucie Cournoyer, Gilles Gagnon, Marie-Gloriose Ingabire, Lucie L'Heureux, Remi Moreau, Clairette Simard and Claudine St-Pierre. Finally, much of the book's production was carried out in the graphics department at HEC. We thank Mireille Donais for her remarkable collaboration on this project.
Montreal, September 1998
App
endi
x
Dea
th c
ause
s in
Que
bec,
Can
ada
and
Uni
ted
Sta
tes
in 1
995
Que
bec
Can
ada
Dea
th
Dea
th
Cau
se
Num
ber
PC
T
Rat
e N
umbe
r P
CT
R
ate
Acc
iden
ts a
nd a
dver
se e
ffec
ts
2,02
1 3.
83
27.5
2 8,
820
4.19
29
.78
Mot
or v
ehic
le a
ccid
ents
83
1 1.
58
11.3
2 3,
231
1.53
10
.91
Air
and
spa
ce t
rans
port
acc
iden
ts
9 0.
02
0.12
96
0.
05
0.32
Can
cers
15
,775
29
.91
214.
82
57,8
10
27.4
3 19
5.20
Hom
icid
es
124
0.24
1.
69
489
0.23
1.
65
HIV
inf
ecti
ons
585
1.11
7.
97
1,76
4 0.
84
5.96
Tot
al d
eath
s 52
,734
21
0,73
3
Est
imat
ed p
opul
atio
n 7,
343,
240
29,6
15,3
25
PCT
:
Perc
enta
ge o
ver
tota
l de
aths
D
eath
Rat
e:
Dea
ths
per
100,
000
indi
vidu
als
Uni
ted
Sta
tes
Num
ber
PC
T
89,7
03
3.88
41,8
17
1.81
961
0.04
537,
969
23.2
7
21,5
77
0.93
42,5
06
1.84
2,31
2,18
0
262,
755,
000
Dea
th
Rat
e
34.1
4
15.9
1
0.37
204.
74
6.21
16.1
8
;.<
;.< S: ~ §5 ~ Z
en
C
;>:l
;I- ~ tIl
App
endi
x
Dea
th c
ause
s in
Que
bec,
Can
ada
and
Uni
ted
Sta
tes
in 1
995
Que
bec
Can
ada
Dea
th
Dea
th
Cau
se
Num
ber
PC
T
Rat
e N
umbe
r P
CT
R
ate
Acc
iden
ts a
nd a
dver
se e
ffec
ts
2,02
1 3.
83
27.5
2 8,
820
4.19
29
.78
Mot
or v
ehic
1e a
ccid
ents
83
1 1.
58
11.3
2 3,
231
1.53
10
.91
Air
and
spa
ce t
rans
port
acc
iden
ts
9 0.
02
0.12
96
0.
05
0.32
Can
cers
15
,775
29
.91
214.
82
57,8
10
27.4
3 19
5.20
Hom
icid
es
124
0.24
1.
69
489
0.23
1.
65
HIV
inf
ecti
ons
585
1.11
7.
97
1,76
4 0.
84
5.96
Tot
al d
eath
s 52
,734
21
0,73
3
Est
imat
ed p
opul
atio
n 7,
343,
240
29,6
15,3
25
PCT
:
Perc
enta
ge o
ver
tota
l dea
ths
Dea
th R
ate:
D
eath
s pe
r 10
0,00
0 in
divi
dual
s
Uni
ted
Sta
tes
Num
ber
PC
T
89,7
03
3.88
41,8
17
1.81
961
0.04
537,
969
23.2
7
21,5
77
0.93
42,5
06
1.84
2,31
2,18
0
262,
755,
000
Dea
th
Rat
e
34.1
4
15.9
1
0.37
204.
74
6.21
16.1
8
;.<
;.< s: ~ §5 ~ z '" c ;>:l
;1- ~ tI1
Definitions
Accidents and adverse effects (AE234-AE263): motor vehicle accidents accidental falls suicides homicides other (fire, drowning, handguns, railway, ... )
Motor vehicle accidents (E81O-E819): traffic accidents (involving collision or without collision) non traffic accidents
Air and space transport accidents (air carriers and general aviation) (E840-E845): accidents to powered aircraft at takeoff and landing accidents to powered aircraft, other and unspecified other specified air transport accidents (parachutist, military, ... )
Cancers (ICD-9 codes: 140-208): malignant neoplasm of lip, oral cavity and pharynx malignant neoplasm of digestive organs and peritoneum malignant neoplasm of respiratory system malignant neoplasm, bone, connective tissue, skin and breast malignant neoplasm of genito-urinary organs malignant neoplasm of other and unspecified sites neoplasm of lymphatic and haematopoietic tissue
Homicide (E960-E969): fight, brawl, rape assault by poisoning assault by hanging and strangulation assault by submersion assault by firearms and explosives assault by cutting and piercing instrument child battering and other maltreatment assault by other and unspecified means late effects of injury purposely inflicted by other person
mv infection (ICD-9 codes: 042-044): with specified conditions causing other specified conditions other
Sources Quebec and Canada:
Statistics Canada 1997, catalogue n° 84-208-XPB: Causes of Death, 1995
XXIX
Statistics Canada 1997, catalogue n° 84-209-XPB: Mortality - Summary List of Causes, 1995
United States: Statistical Abstract of the United States, 1997
The National Data Book U.S. Department of Commerce Economic and Statistics Administration Bureau of Census
Detinitions Accidents and adverse effects (AE234-AE263):
motor vehic1e accidents accidental falIs suicides homicides other (fire, drowning, handguns, railway, ... )
Motor vehic1e accidents (E81O-E819): traffic accidents (involving colIi sion or without collision) non traffic accidents
Air and space transport accidents (air carriers and general aviation) (E840-E845): accidents to powered aircraft at takeoff and landing accidents to powered aircraft, other and unspecified other specified air transport accidents (parachutist, military, ... )
Cancers (ICD-9 codes: 140-208): malignant neoplasm of lip, oral cavity and pharynx malignant neoplasm of digestive organs and peritoneum malignant neoplasm of respiratory system malignant neoplasm, bone, connective tissue, skin and breast malignant neoplasm of genito-urinary organs malignant neoplasm of other and unspecified sites neoplasm of lymphatic and haematopoietic tissue
Homicide (E960-E969): fight, brawl, rape assault by poisoning assault by hanging and strangulation assault by submersion assault by firearms and explosives assault by cutting and piercing instrument child battering and other maltreatment assauIt by other and unspecified means late effects of injury purposely inflicted by other person
HIV infection (ICD-9 codes: 042-044): with specified conditions causing other specified conditions other
Sources Quebec and Canada:
Statistics Canada 1997, catalogue n° 84-208-XPB: Causes of Death, 1995
XXIX
Statistics Canada 1997, catalogue n° 84-209-XPB: Mortality - Summary List of Causes, 1995
United States: Statistical Abstract of the United States, 1997
The National Data Book U.S. Department of Commerce Economic and Statistics Administration Bureau of Census
top related