Road safety for fleets of vehicles has been neglected in the insurance literature, mainly because... more Road safety for fleets of vehicles has been neglected in the insurance literature, mainly because appropriate data is not available. This paper makes a threefold contribution: 1) Produce statistics on current fleets' road safety offences using a panel of 20 years of data on truck fleets; 2) relate these offences to fleets' accidents; and 3) identify and classify the riskiest fleets for insurance ratemaking based on past experience in managing road safety. Our results show a substantial heterogeneity between fleets in terms of road safety.
We contribute to the growing literature on moral hazard by offering empirical evidence of the eff... more We contribute to the growing literature on moral hazard by offering empirical evidence of the effectiveness of insurance pricing incentives at improving road safety by comparing the claim frequency following a regulatory reform introduced in a pilot city in China with the experience of another city unaffected by the reform. By using the methodology of difference-indifferences , we find that improving insurance pricing on past claims and on traffic violations with full industry commitment reduces moral hazard and insured drivers' claim frequency by 12%. The treatment effects are, however, heterogeneous with respect to insured drivers' wealth and past claims history.
The author would like to thank Marie-Gloriose Ingabire for her help with bibliographical research... more The author would like to thank Marie-Gloriose Ingabire for her help with bibliographical research, and FQRSC-Quebec and SSHRC-Canada for their financial support. He would also like to acknowledge the researchers who helped him develop several of the ideas on the subject over the years:
We propose a new parametric model for the modelling and estimation of event distributions for ind... more We propose a new parametric model for the modelling and estimation of event distributions for individuals in different firms. The analysis uses panel data and takes into account individual and firm effects in a non-linear model. Non-observable factors are treated as random effects. In our application, the distribution of accidents is affected by observable and non-observable factors from vehicles, drivers, and fleets of vehicles. Observable and unobservable factors are significant to explain road accidents, which means that insurance pricing should take into account all these factors. A fixed effects model is also estimated to test the consistency of the random effects model.
We survey recent developments in the economic analysis of insurance fraud. The paper first sets o... more We survey recent developments in the economic analysis of insurance fraud. The paper first sets out the two main approaches to insurance fraud that have been developped in the literature, namely the costly state verification and the costly state falsification. Under costly state verification, the insurer can verify claims at some cost. Claims' verification may be deterministic or random, and it can be conditioned on fraud signals perceived by insurers. Under costly state falsification, the policyholder expends resources for the building-up of his or her claim not to be detected. We also consider the effects of adverse selection, in a context where insurers cannot distinguish honest policyholders from potential defrauders, as well as the consequences of credibility constraints on anti-fraud policies. Finally, we focus attention on the risk of collusion between policyholders and insurance agents or service providers.
Nous proposons un modèle paramétrique de tarification de l’assurance de véhicules routiers appart... more Nous proposons un modèle paramétrique de tarification de l’assurance de véhicules routiers appartenant à une flotte. Les tables de primes qui y sont présentées tiennent compte des accidents passés des véhicules, des caractéristiques observables des véhicules et des flottes et des infractions au Code de la sécurité routière des conducteurs et des transporteurs. De plus, les primes sont ajustées en fonction des accidents accumulés par les flottes dans le temps. Il s’agit d’un modèle qui prend directement en compte des changements explicites des différentes composantes des probabilités d’accidents. Il représente une extension aux modèles d’assurance automobile de type bonus-malus pour les primes individuelles (Lemaire, 1985 ; Dionne et Vanasse, 1989 et 1992 ; Pinquet, 1997 et 1998 ; Frangos et Vrontos, 2001 ; Purcaru et Denuit, 2003). L’extension ajoute un effet flotte à l’effet véhicule pour tenir compte des caractéristiques ou des actions non observables des transporteurs sur les tau...
In this survey we present some of the more signi…cant results in the literature on adverse select... more In this survey we present some of the more signi…cant results in the literature on adverse selection in insurance markets. Sections 1 and 2 introduce the subject and Section 3 discusses the monopoly model developed by Stiglitz (1977) for the case of single-period contracts extended by many authors to the multi-period case. The introduction of multi-period contracts raises many issues that are discussed in detail; time horizon, discounting, commitment of the parties, contract renegotiation and accidents underreporting. Section 4 covers the literature on competitive contracts. The analysis is more complicated because insurance companies must take into account competitive pressures when they set incentive contracts. As pointed out by Rothschild and Stiglitz (1976), there is not necessarily a Cournot-Nash equilibrium in the presence of adverse selection. However, market equilibrium can be sustained when principals anticipate competitive reactions to their behavior or when they adopt strategies that di¤er from the pure Nash strategy. Multi-period contracting is discussed. We show that di¤erent predictions on the evolution of insurer pro…ts over time can be obtained from di¤erent assumptions concerning the sharing of information between insurers about individual's choice of contracts and accident experience. The roles of commitment and renegotiation between the parties to the contract are important. Section 5 introduces models that consider moral hazard and adverse selection simultaneously and Section 6 covers adverse selection when people can choose their risk status. Section 7 discusses many extensions to the basic models such as risk categorization, multidimensional adverse selection, symmetric imperfect information, reversed or double-sided adverse selection, principals more informed than agents, uberrima …des and participating contracts.
This research was financed by the Fonds pour la formation de chercheurs et l'aide à la recherche ... more This research was financed by the Fonds pour la formation de chercheurs et l'aide à la recherche (F.C.A.R., Quebec), the Société d'assurance automobile du Québec (SAAQ), the Ministère des transports du Québec (M.T.Q.), CREST and the Fédération Française des Sociétés d'Assurances (FFSA).
Traditional insurance contracts do not offer protection against the replacement value of a vehicl... more Traditional insurance contracts do not offer protection against the replacement value of a vehicle. A replacement cost endorsement gives the opportunity to get a new vehicle in the case of a total theft or in the case of total destruction of the car in a road accident. This type of protection was introduced in Canada in the late 1980's. It is also offered in France and many insurers in the United States are going to move in that direction. We show that holders of car insurance policies with a replacement cost endorsement have a higher probability of theft near the end of this additional protection (usually 24 months following the acquisition of a new car). Our tests indicate that this result is a form of ex-post moral hazard or opportunistic insurance fraud.
L’économétrie des transports et de l’assurance, 1997
Le but de cette recherche était d’évaluer l’effet du changement de tarification de 1992 sur la sé... more Le but de cette recherche était d’évaluer l’effet du changement de tarification de 1992 sur la sécurité routière au Québec. Nos résultats indiquent que le changement de tarification a réduit les nombres d’infractions et les nombres d’accidents, deux variables qui mesurent indirectement la non-prévention routière. De plus, nos résultats indiquent que le nombre de points d’inaptitude accumulés au cours d’une période de deux ans est un bon prédicteur du nombre d’accidents de la période suivante de deux ans ce qui supporte la politique de tarification de la SAAQ. En effet, cette politique en plus d’inciter plus de prudence, fait payer des contributions d’assurance proportionnelles aux risques individuels. En d’autres termes, le changement de 1992 a réintroduit une tarification des risques plus équitable au sens actuariel en faisant payer aux risques élevés des contributions d’assurance plus élevées. Ces résultats ont été obtenus de l’estimation des paramètres de la loi de distribution b...
Risk classification refers to the use of observable characteristics by insurers to group individu... more Risk classification refers to the use of observable characteristics by insurers to group individuals with similar expected claims, to compute the corresponding premiums, and thereby to reduce asymmetric information. Permitting risk classification may reduce informational asymmetry-induced adverse selection and improve insurance market efficiency. It may also have undesirable equity consequences and undermine the implicit insurance against reclassification risk, which legislated restrictions on risk classification could provide. We use a canonical insurance market screening model to survey and to extend the risk classification literature. We provide a unified framework for analysing the economic consequences of legalised vs banned risk classification, both in static-information environments and in environments in which additional information can be learned, by either side of the market, through potentially costly tests.
Nous abordons la difficile question de la mesure empirique des effets des problèmes d’information... more Nous abordons la difficile question de la mesure empirique des effets des problèmes d’information sur l’allocation des ressources. Deux problèmes retiennent notre attention; le risque moral et l’antisélection. Une conclusion, acceptée par la plupart des auteurs, est que les problèmes d’information créent des distorsions importantes dans l’économie. Mais nous pouvons vérifier, dans certains marchés, que des mécanismes efficaces ont été mis en place pour réduire ces distorsions et même éliminer, à la marge, des problèmes résiduels d’information. Cette conclusion semble plus forte pour l’antisélection que pour le risque moral. Une explication est le fait que l’antisélection concerne des caractéristiques exogènes, alors que le risque moral est expliqué par des actions endogènes qui peuvent être modifiées en tout temps.
Risk classification refers to the use of observable characteristics by insurers to group individu... more Risk classification refers to the use of observable characteristics by insurers to group individuals with similar expected claims, compute the corresponding premiums, and thereby reduce asymmetric information. With perfect risk classification, premiums fully reflect the expected cost associated with each class of risk characteristics and yield efficient outcomes. In the health sector, risk classification is also subject to concerns about social equity and potential discrimination. We present an analytical framework that illustrates the potential trade-off between efficient insurance provision and social equity. We also review empirical studies on risk classification and residual asymmetric information that inform this trade-off.
In this paper, we estimate the impact of introducing a bonus-malus system on the probability of h... more In this paper, we estimate the impact of introducing a bonus-malus system on the probability of having automobile accidents, taking into account contract duration or the client mobility between insurers. We show that the new incentive scheme reduces accident rates of all policyholders when contract duration is taken into account, but does not a¤ect accident rates of movers that shirk the imposed incentive e¤ects of the new insurance pricing scheme.
... (2005)1. Beaulieu, M.-C., Dufour, J.-M. and Khalaf, L. 2005. ... if (R 1 , R 2 ) is positivel... more ... (2005)1. Beaulieu, M.-C., Dufour, J.-M. and Khalaf, L. 2005. ... if (R 1 , R 2 ) is positively quadrant dependent, then both Cov(R 1 , R 2 ) and ... Franklin Allen, Marie-Claude Beaulieu, Oussama Chakroun, Jean-Claude Cosset, Pascal François, Julien Hugonnier, Nadia Ouertani, Jean ...
This article makes a bridge between the theory of optimal auditing and the scoring methodology in... more This article makes a bridge between the theory of optimal auditing and the scoring methodology in an asymmetric information setting. Our application is meant for insurance claims fraud, but it can be applied to many other activities that use the scoring approach. Fraud signals are classified based on the degree to which they reveal an increasing probability of fraud. We show that the optimal auditing strategy takes the form of a “red flags strategy,” which consists in referring claims to a special investigative unit (SIU) when certain fraud indicators are observed. The auditing policy acts as a deterrence device, and we explain why it requires the commitment of the insurer and how it should affect the incentives of SIU staffs. The characterization of the optimal auditing strategy is robust to some degree of signal manipulation by defrauders as well as to the imperfect information of defrauders about the audit frequency. The model is calibrated with data from a large European insuran...
Financial institutions face various cyclical risks, but very few studies have analyzed the cyclic... more Financial institutions face various cyclical risks, but very few studies have analyzed the cyclicality of operational risk. External fraud is an important operational risk faced by insurers. In this research, we analyze the empirical relationship between insurance fraud and business cycle and we concentrate our study on two insurance contracts that may create an incentive to defraud. We find that residual insurance fraud exists both in the contract with replacement cost endorsement and the contract with no-deductible endorsement in the Taiwan automobile theft insurance market. These results are consistent with previous literature on the relationship between fraud activity and non-optimal insurance contracting. We also show that the severity of insurance fraud is countercyclical. Fraud is stimulated during periods of recession and mitigated during periods of expansion. Although this last result seems intuitive, our contribution is the first to measure its significance.
Brassard, and Lyne Vézina for their collaboration at various stages of this project. Stéphane Mes... more Brassard, and Lyne Vézina for their collaboration at various stages of this project. Stéphane Messier made an excellent contribution to the preparation and the management of the survey and Claire Boisvert improved significantly the presentation of the original manuscript. We thank Michèle Cohen, the editor, and an anonymous referee for very useful comments.
Journal of Risk <html_ent glyph="@amp;" ascii="&amp;"/> Insurance, 2005
The objective of this study is to assess empirically what impact introduction of the bonusmalus s... more The objective of this study is to assess empirically what impact introduction of the bonusmalus system has had on road safety in Tunisia. The results of the Tunisian experiment are now of particular importance since, during the last decade, many European countries decided to eliminate their bonus-malus scheme. These results indicate that the bonusmalus system reduced the probability of reported accident for good risks but had no effect on that of bad risks. Moreover, the overall effect of the reform on reported accidents rates is not statistically significant. This finding is explained by the fact that bad risks can switch to another insurer so as to escape the incentive effect imposed by the new rating policy. Many control variables are statistically significant in explaining the number of reported accidents: the vehicle's horsepower, the policyholder's place of residence, exits from the insurer portfolio, and the coverages for which policyholders are underwritten. The coefficient of the predicted exit variable is positive in explaining the number of accidents. This indicates that policyholders who switch company are higher risks. The final results were obtained by introducing random individual-specific effects to make joint estimates of the accident and selection equations.
Road safety for fleets of vehicles has been neglected in the insurance literature, mainly because... more Road safety for fleets of vehicles has been neglected in the insurance literature, mainly because appropriate data is not available. This paper makes a threefold contribution: 1) Produce statistics on current fleets' road safety offences using a panel of 20 years of data on truck fleets; 2) relate these offences to fleets' accidents; and 3) identify and classify the riskiest fleets for insurance ratemaking based on past experience in managing road safety. Our results show a substantial heterogeneity between fleets in terms of road safety.
We contribute to the growing literature on moral hazard by offering empirical evidence of the eff... more We contribute to the growing literature on moral hazard by offering empirical evidence of the effectiveness of insurance pricing incentives at improving road safety by comparing the claim frequency following a regulatory reform introduced in a pilot city in China with the experience of another city unaffected by the reform. By using the methodology of difference-indifferences , we find that improving insurance pricing on past claims and on traffic violations with full industry commitment reduces moral hazard and insured drivers' claim frequency by 12%. The treatment effects are, however, heterogeneous with respect to insured drivers' wealth and past claims history.
The author would like to thank Marie-Gloriose Ingabire for her help with bibliographical research... more The author would like to thank Marie-Gloriose Ingabire for her help with bibliographical research, and FQRSC-Quebec and SSHRC-Canada for their financial support. He would also like to acknowledge the researchers who helped him develop several of the ideas on the subject over the years:
We propose a new parametric model for the modelling and estimation of event distributions for ind... more We propose a new parametric model for the modelling and estimation of event distributions for individuals in different firms. The analysis uses panel data and takes into account individual and firm effects in a non-linear model. Non-observable factors are treated as random effects. In our application, the distribution of accidents is affected by observable and non-observable factors from vehicles, drivers, and fleets of vehicles. Observable and unobservable factors are significant to explain road accidents, which means that insurance pricing should take into account all these factors. A fixed effects model is also estimated to test the consistency of the random effects model.
We survey recent developments in the economic analysis of insurance fraud. The paper first sets o... more We survey recent developments in the economic analysis of insurance fraud. The paper first sets out the two main approaches to insurance fraud that have been developped in the literature, namely the costly state verification and the costly state falsification. Under costly state verification, the insurer can verify claims at some cost. Claims' verification may be deterministic or random, and it can be conditioned on fraud signals perceived by insurers. Under costly state falsification, the policyholder expends resources for the building-up of his or her claim not to be detected. We also consider the effects of adverse selection, in a context where insurers cannot distinguish honest policyholders from potential defrauders, as well as the consequences of credibility constraints on anti-fraud policies. Finally, we focus attention on the risk of collusion between policyholders and insurance agents or service providers.
Nous proposons un modèle paramétrique de tarification de l’assurance de véhicules routiers appart... more Nous proposons un modèle paramétrique de tarification de l’assurance de véhicules routiers appartenant à une flotte. Les tables de primes qui y sont présentées tiennent compte des accidents passés des véhicules, des caractéristiques observables des véhicules et des flottes et des infractions au Code de la sécurité routière des conducteurs et des transporteurs. De plus, les primes sont ajustées en fonction des accidents accumulés par les flottes dans le temps. Il s’agit d’un modèle qui prend directement en compte des changements explicites des différentes composantes des probabilités d’accidents. Il représente une extension aux modèles d’assurance automobile de type bonus-malus pour les primes individuelles (Lemaire, 1985 ; Dionne et Vanasse, 1989 et 1992 ; Pinquet, 1997 et 1998 ; Frangos et Vrontos, 2001 ; Purcaru et Denuit, 2003). L’extension ajoute un effet flotte à l’effet véhicule pour tenir compte des caractéristiques ou des actions non observables des transporteurs sur les tau...
In this survey we present some of the more signi…cant results in the literature on adverse select... more In this survey we present some of the more signi…cant results in the literature on adverse selection in insurance markets. Sections 1 and 2 introduce the subject and Section 3 discusses the monopoly model developed by Stiglitz (1977) for the case of single-period contracts extended by many authors to the multi-period case. The introduction of multi-period contracts raises many issues that are discussed in detail; time horizon, discounting, commitment of the parties, contract renegotiation and accidents underreporting. Section 4 covers the literature on competitive contracts. The analysis is more complicated because insurance companies must take into account competitive pressures when they set incentive contracts. As pointed out by Rothschild and Stiglitz (1976), there is not necessarily a Cournot-Nash equilibrium in the presence of adverse selection. However, market equilibrium can be sustained when principals anticipate competitive reactions to their behavior or when they adopt strategies that di¤er from the pure Nash strategy. Multi-period contracting is discussed. We show that di¤erent predictions on the evolution of insurer pro…ts over time can be obtained from di¤erent assumptions concerning the sharing of information between insurers about individual's choice of contracts and accident experience. The roles of commitment and renegotiation between the parties to the contract are important. Section 5 introduces models that consider moral hazard and adverse selection simultaneously and Section 6 covers adverse selection when people can choose their risk status. Section 7 discusses many extensions to the basic models such as risk categorization, multidimensional adverse selection, symmetric imperfect information, reversed or double-sided adverse selection, principals more informed than agents, uberrima …des and participating contracts.
This research was financed by the Fonds pour la formation de chercheurs et l'aide à la recherche ... more This research was financed by the Fonds pour la formation de chercheurs et l'aide à la recherche (F.C.A.R., Quebec), the Société d'assurance automobile du Québec (SAAQ), the Ministère des transports du Québec (M.T.Q.), CREST and the Fédération Française des Sociétés d'Assurances (FFSA).
Traditional insurance contracts do not offer protection against the replacement value of a vehicl... more Traditional insurance contracts do not offer protection against the replacement value of a vehicle. A replacement cost endorsement gives the opportunity to get a new vehicle in the case of a total theft or in the case of total destruction of the car in a road accident. This type of protection was introduced in Canada in the late 1980's. It is also offered in France and many insurers in the United States are going to move in that direction. We show that holders of car insurance policies with a replacement cost endorsement have a higher probability of theft near the end of this additional protection (usually 24 months following the acquisition of a new car). Our tests indicate that this result is a form of ex-post moral hazard or opportunistic insurance fraud.
L’économétrie des transports et de l’assurance, 1997
Le but de cette recherche était d’évaluer l’effet du changement de tarification de 1992 sur la sé... more Le but de cette recherche était d’évaluer l’effet du changement de tarification de 1992 sur la sécurité routière au Québec. Nos résultats indiquent que le changement de tarification a réduit les nombres d’infractions et les nombres d’accidents, deux variables qui mesurent indirectement la non-prévention routière. De plus, nos résultats indiquent que le nombre de points d’inaptitude accumulés au cours d’une période de deux ans est un bon prédicteur du nombre d’accidents de la période suivante de deux ans ce qui supporte la politique de tarification de la SAAQ. En effet, cette politique en plus d’inciter plus de prudence, fait payer des contributions d’assurance proportionnelles aux risques individuels. En d’autres termes, le changement de 1992 a réintroduit une tarification des risques plus équitable au sens actuariel en faisant payer aux risques élevés des contributions d’assurance plus élevées. Ces résultats ont été obtenus de l’estimation des paramètres de la loi de distribution b...
Risk classification refers to the use of observable characteristics by insurers to group individu... more Risk classification refers to the use of observable characteristics by insurers to group individuals with similar expected claims, to compute the corresponding premiums, and thereby to reduce asymmetric information. Permitting risk classification may reduce informational asymmetry-induced adverse selection and improve insurance market efficiency. It may also have undesirable equity consequences and undermine the implicit insurance against reclassification risk, which legislated restrictions on risk classification could provide. We use a canonical insurance market screening model to survey and to extend the risk classification literature. We provide a unified framework for analysing the economic consequences of legalised vs banned risk classification, both in static-information environments and in environments in which additional information can be learned, by either side of the market, through potentially costly tests.
Nous abordons la difficile question de la mesure empirique des effets des problèmes d’information... more Nous abordons la difficile question de la mesure empirique des effets des problèmes d’information sur l’allocation des ressources. Deux problèmes retiennent notre attention; le risque moral et l’antisélection. Une conclusion, acceptée par la plupart des auteurs, est que les problèmes d’information créent des distorsions importantes dans l’économie. Mais nous pouvons vérifier, dans certains marchés, que des mécanismes efficaces ont été mis en place pour réduire ces distorsions et même éliminer, à la marge, des problèmes résiduels d’information. Cette conclusion semble plus forte pour l’antisélection que pour le risque moral. Une explication est le fait que l’antisélection concerne des caractéristiques exogènes, alors que le risque moral est expliqué par des actions endogènes qui peuvent être modifiées en tout temps.
Risk classification refers to the use of observable characteristics by insurers to group individu... more Risk classification refers to the use of observable characteristics by insurers to group individuals with similar expected claims, compute the corresponding premiums, and thereby reduce asymmetric information. With perfect risk classification, premiums fully reflect the expected cost associated with each class of risk characteristics and yield efficient outcomes. In the health sector, risk classification is also subject to concerns about social equity and potential discrimination. We present an analytical framework that illustrates the potential trade-off between efficient insurance provision and social equity. We also review empirical studies on risk classification and residual asymmetric information that inform this trade-off.
In this paper, we estimate the impact of introducing a bonus-malus system on the probability of h... more In this paper, we estimate the impact of introducing a bonus-malus system on the probability of having automobile accidents, taking into account contract duration or the client mobility between insurers. We show that the new incentive scheme reduces accident rates of all policyholders when contract duration is taken into account, but does not a¤ect accident rates of movers that shirk the imposed incentive e¤ects of the new insurance pricing scheme.
... (2005)1. Beaulieu, M.-C., Dufour, J.-M. and Khalaf, L. 2005. ... if (R 1 , R 2 ) is positivel... more ... (2005)1. Beaulieu, M.-C., Dufour, J.-M. and Khalaf, L. 2005. ... if (R 1 , R 2 ) is positively quadrant dependent, then both Cov(R 1 , R 2 ) and ... Franklin Allen, Marie-Claude Beaulieu, Oussama Chakroun, Jean-Claude Cosset, Pascal François, Julien Hugonnier, Nadia Ouertani, Jean ...
This article makes a bridge between the theory of optimal auditing and the scoring methodology in... more This article makes a bridge between the theory of optimal auditing and the scoring methodology in an asymmetric information setting. Our application is meant for insurance claims fraud, but it can be applied to many other activities that use the scoring approach. Fraud signals are classified based on the degree to which they reveal an increasing probability of fraud. We show that the optimal auditing strategy takes the form of a “red flags strategy,” which consists in referring claims to a special investigative unit (SIU) when certain fraud indicators are observed. The auditing policy acts as a deterrence device, and we explain why it requires the commitment of the insurer and how it should affect the incentives of SIU staffs. The characterization of the optimal auditing strategy is robust to some degree of signal manipulation by defrauders as well as to the imperfect information of defrauders about the audit frequency. The model is calibrated with data from a large European insuran...
Financial institutions face various cyclical risks, but very few studies have analyzed the cyclic... more Financial institutions face various cyclical risks, but very few studies have analyzed the cyclicality of operational risk. External fraud is an important operational risk faced by insurers. In this research, we analyze the empirical relationship between insurance fraud and business cycle and we concentrate our study on two insurance contracts that may create an incentive to defraud. We find that residual insurance fraud exists both in the contract with replacement cost endorsement and the contract with no-deductible endorsement in the Taiwan automobile theft insurance market. These results are consistent with previous literature on the relationship between fraud activity and non-optimal insurance contracting. We also show that the severity of insurance fraud is countercyclical. Fraud is stimulated during periods of recession and mitigated during periods of expansion. Although this last result seems intuitive, our contribution is the first to measure its significance.
Brassard, and Lyne Vézina for their collaboration at various stages of this project. Stéphane Mes... more Brassard, and Lyne Vézina for their collaboration at various stages of this project. Stéphane Messier made an excellent contribution to the preparation and the management of the survey and Claire Boisvert improved significantly the presentation of the original manuscript. We thank Michèle Cohen, the editor, and an anonymous referee for very useful comments.
Journal of Risk <html_ent glyph="@amp;" ascii="&amp;"/> Insurance, 2005
The objective of this study is to assess empirically what impact introduction of the bonusmalus s... more The objective of this study is to assess empirically what impact introduction of the bonusmalus system has had on road safety in Tunisia. The results of the Tunisian experiment are now of particular importance since, during the last decade, many European countries decided to eliminate their bonus-malus scheme. These results indicate that the bonusmalus system reduced the probability of reported accident for good risks but had no effect on that of bad risks. Moreover, the overall effect of the reform on reported accidents rates is not statistically significant. This finding is explained by the fact that bad risks can switch to another insurer so as to escape the incentive effect imposed by the new rating policy. Many control variables are statistically significant in explaining the number of reported accidents: the vehicle's horsepower, the policyholder's place of residence, exits from the insurer portfolio, and the coverages for which policyholders are underwritten. The coefficient of the predicted exit variable is positive in explaining the number of accidents. This indicates that policyholders who switch company are higher risks. The final results were obtained by introducing random individual-specific effects to make joint estimates of the accident and selection equations.
Uploads
Papers by georges dionne