How an interconnected ecosystem can address this pernicious problem
By Srivathsan Karanai Margan
In insurance, moral hazard is the idea that a party that is protected from risk will behave differently than they would if they lacked that protection. It has always been difficult to track moral hazard and initiate preventive mitigation measures. Insurers have traditionally adopted risk control techniques and restrictive practices to mitigate the impact of moral hazard. But in an interconnected world, insurers will be able to accurately track moral hazard in real time, potentially resulting in changing moral hazard from being a lagging indicator into a cutting-edge leading indicator.
New technologies, including the “internet of things” (IOT), Big Data, cloud computing, and artificial intelligence (AI), are disrupting businesses and serving as a fertile platform for many new business models to evolve. IOT is especially animating many of the devices around us, feeding data pipelines with real-time data and creating a connected ecosystem. As it permeates, this connectedness is causing fundamental changes on existing businesses—and obliterating a few in the process.
These technologies are also disrupting the insurance industry, allowing businesses to deliver a new, connected insurance landscape. Many insurance and insurtech companies are introducing new product and service offerings that continuously monitor the condition of that which is being insured, its usage, ongoing maintenance, and behavior of the customer. Insurers provide preventive and care services along with various nudge elements for achieving positive behavioral changes to reduce risk levels. All of this is bringing about an epochal shift in insurance, where the role of insurance is evolving from risk transfer to risk prevention.
Considering the benefits that connected insurance can deliver in personal lines (auto, home, life, and health), adoption has been surprisingly slow and restricted to a few innovators and geographies. This low conversion rate could be attributed to two important factors:
- Privacy concerns—The infrastructure of connected insurance is built on the seamless customer data that flows from the devices. The data privacy and security concerns of the customers, and the fear of surveillance or bad actors accessing customers’ data, could be an important limiting factor.
- Insurers are fringe players—IOT device companies are the hosts of the connected ecosystem and hence its most important stakeholders. Today, insurance is just a financial link in the ecosystem to incentivize customers—and hence insurers are mere peripheral players. They are not at the moment in a position to influence the course of the connected insurance in a meaningful way.
But I posit that these are mere teething problems that will dissipate as technologies and the interconnected ecosystem mature. The customer fear to adopt is quite natural in the starting stages of any new technology. For many, these apprehensions will disappear as they realize that the benefits obtained by sharing data are sizeable and there is transparency in the way data is going to be used—with whom the data is likely to be shared and why. Connected insurance is still in the initial stages of growth, with various engagement models and offerings evolving. As these models and products emerge, insurers may operate new customer assistance activities and play an important role in the interconnected ecosystem.
There is little doubt that some large proportion of the insurance of the future will be connected. A new order is bound to emerge in which insurance will be related to more positive occurrences, such as risk prevention and healthy living, rather than something negative. Connected insurance can be structured to benefit customers and insurers alike: While risk prevention, personalization of products and services, and focus on well-being will benefit customers, insurers will benefit from reduced claim losses and contextual intelligence with respect to the usage and behavior of customers.
Moral Hazard as We Know It
The occurrence of an insured risk and the extent of loss depends on perils and hazards. A peril is defined as the probable cause that exposes an insured person or object to the risk of damage, injury, or loss against which an insurance cover is purchased. Hazard is defined as a condition that has the potential to increase the frequency and/or severity of the loss arising from the peril. Specifically, behavioral hazard arises when a person acts either dishonestly or indifferently.
Moral hazard arises when individuals believe the benefit gained by violating the terms of their contract to be more valuable than the probability and the costs of being caught, or when they tend to behave differently because they do not have to bother about the cost of a loss. Without complete knowledge about the behavior, insurers accept risks or settle claims that they would have otherwise declined or considered on different terms.
Behavioral hazard is classified in two ways depending either on the intent of customer or the behavioral changes relating to post-purchase of insurance and post-occurrence of a covered event. Regarding customer intent, behavioral hazard is further subclassified as moral hazard and morale hazard. Moral hazard occurs when a customer intentionally suppresses facts material to a contract, acts dishonestly, stages an accident, fakes an injury, or willfully causes, fabricates, or exaggerates a loss. Morale hazard, on the other hand, results when a customer is careless, lazy, or indifferent and fails to take adequate care of that for which insurance is held or fails to initiate steps to prevent or mitigate losses.
Related to behavioral changes, behavior hazard is further subclassified as ex-ante moral hazard (EAMH) and ex-post moral hazard (EPMH). EAMH refers to a harmful change in attitude after purchase of insurance but before ny incidents occur such that there is a reduced preventive effort. For example, speeding while driving a car or failing to lock a vehicle or house. Examples of EPMH could be either harmless or harmful. The incidence of EPMH could be as banal as increasing usage of a service just because it is covered, or it could be illicit, such as inflating claim bills, filing false claims, and avoiding reporting losses to qualify for lower-risk incentives.
Comparing the two ways of classification, moral hazard could be either EAMH or EPMH, whereas morale hazard is only EPMH.
Traditional Ways of Controlling Moral Hazard
In an unconnected world of insurance, insurers cannot discern violations that could occur anytime during the entire course of an insurance contract. In most of the insurance contracts, moral hazard assessment that is done at inception is presumed to continue throughout the contract period. Property and casualty insurance contracts are for shorter terms and insurers get to reassess the risk periodically. Life insurance contracts, on the other hand, are in force for longer periods and are automatically renewable on payment of annual premiums without being re-underwritten.
Insurers may depend on the following sources to get data to deduce, presume, and predict moral hazard to construct models, classify risks, and define the terms of the contract:
Customers are required to disclose all the information that is material to the contract, both at inception and during the course of the contract. They are supposed to adhere to the principles of loss minimization by taking all necessary steps to reduce harm or damage. However, the prevalent information asymmetry creates the space for moral hazard to arise.
To curb the loss from moral hazard, insurers apply risk control techniques such as loss prevention and loss reduction. At the time of underwriting the policies, these techniques help insurers to restrict the extent of exposure and claim liabilities. To forestall moral hazard, insurers use loss prevention practices such as providing coverage only if a customer adheres to certain security norms (installation of heating systems and automatic sprinklers, for example) or goes through scrutiny procedures (a nicotine test to identify smoking status).
Insurers apply restrictive provisions, such as modifying coverage type, classifying in higher risk category, charging a higher premium, limiting exposure, or imposing a waiting period for coverage commencement, to reduce the exposure to risk and scale down moral hazard. For loss reduction at the time of settling claims, insurers apply some pre-imposed exclusions, clauses, and riders. Insurers also insist customers share a portion of the claim amount (deductibles, coinsurance, and copay) so that the unease in bearing a portion of the loss forestalls any moral hazard violation.
There are also a few restrictive provisions that are executed during the term of the policy. Increase in hazard provisions allows insurers to limit or suspend all or part of the risk when there is an increase in hazard and restore coverage when it subsides—property/fire insurance, for example. The protective safeguards provision allows suspension of coverage when a protective device is not functioning and restore the coverage when the safeguard is restored, as is the case with camera and musical instrument dealers insurance. The inspection and surveys provision permits commercial insurers to inspect the premises of the insured during the policy period for boiler and machinery insurance or workers’ compensation policies to conduct loss control inspection. On discovering any previously unknown hazard, the insurer may decide to deny a claim, increase the premium, suspend or cancel risk, or not renew the contract.
Revisiting Control With Continuous Monitoring
The connected IOT devices sense, respond, communicate, and initiate action. These devices continuously collect and feed the data pipeline with information pertaining to the parameter tracked, action initiated, its own condition, and that of the environment. Continuous real-time monitoring achieved with these devices enables insurers not only to intervene for proactive risk prevention, but also to assess customer behavior.
IOT and its fusion with AI will create the capability to consume the data deluge, analyze, interpret, and provide deeper actionable insights to individuals for preventing or reducing the impact of risk. In the connected insurance landscape created by this union, many of the potential losses caused by usage and behavior that were traditionally believed to be uncontrollable will be brought into the controllable realm. Consequently, the philosophy of insurance will shift from risk transfer and remediation to risk monitoring, prevention, and protection.
Insurers are increasingly empowered by real-time data that they never before had access to. By monitoring and analyzing this data, a perilous condition can be identified before it eventuates, and proactive risk prevention measures can be adopted to reduce the claim losses. It is also possible for insurers to ascertain how a sensor is used and maintained—and more important, how a customer or support system responds to an actionable alert.
From the connected insurance perspective, insurers may need to redefine the product and service offerings and the continuous monitoring standards. They may have to guide customers regarding the type of smart devices, including wearables, that have to be installed or used, and the preventive and support services they will provide with their ecosystem partners. Under the protective safeguards provision, insurers may prescribe customers to follow certain standards and procedures for using the devices, such as how to maintain, periodic servicing, upgrades to be done, and how to respond to actionable insights. Insurers will use nudge elements for achieving behavioral reinforcement for risk prevention and well-being. Insureds will find themselves being rewarded or penalized based on their behavior.
The possibility of moral hazard leading to a claim may be significantly mitigated when a customer follows the standards mandated, allows the usage and behavior to be monitored, and adapts behavior to react to reinforcement. This may cause a substantial decrease in the claim losses that are generated because of controllable behavior-related factors, and the claims that arise are more likely to be genuine claims resulting from pure risk events. The claims that arise in spite of strict adherence to procedures may be settled on a fast track without any restrictions. Insureds will benefit as insurers relax the restrictive risk management techniques that they have traditionally applied as a proxy for containing moral hazard. For example, in case of genuine claims, deductibles may be lowered or eliminated, and a claim won’t necessarily result in a rate hike. If no claims arise during the policy period, customers may get higher premium discounts.
Customers who do not follow the procedures may be proactively nudged to rectify their behavior. Failing to take remedial steps may be considered as not following the principles of loss minimization and hence an incidence of increase in hazard. Insurers may enforce the risk control measures that exist today in an effective way. The risk may either be immediately suspended, exclusions may be imposed to control the exposure, or the risk cover may be restricted to only fortuitous losses. Repeated inaction or inappropriate action to actionable alerts will be considered a mistake, and newly defined mistake exclusion provisions may be invoked to deny claims. For those cases in which there were instances of moral hazard aberrations that did not result in a claim, the violations may be treated as intentional and appropriately weighed while considering the policy for renewal.
The Path and the Future
Considering the possibility in the reduction of loss frequency and severity, insurers are strong advocates of connected insurance. For example, insurance companies such as Progressive, Allstate, Nationwide (auto); State Farm, Liberty Mutual, American Family Insurance (home); John Hancock, Sureify, Phoenix (life); and Cigna, Humana, UnitedHealth, Oscar (health), are enticing customers with tangible financial rewards to embrace connected insurance. These insurers are offering discounts in the cost of devices or/and premiums for installing them (home) or for displaying positive behavioral traits (auto, life, and health). When the measured behavior is less than optimal, insurers are merely reducing or denying the additional financial rewards. Currently, these programs work on a reward/no reward basis, with no penalty for bad behavior. As the growth is still in the early stages, insurers are not following the “bonus-malus” system, which is to reward good behavior and impose restrictions for a negative one. However, as the ecosystem matures, insurers may start imposing the risk-control practices more effectively.
As of now, the growth of the connected ecosystem may be limited because of concerns with respect to data sharing. Distributed ledger technology (DLT) provides an immutable and append-only ledger, and with the evolution of multiple platforms such as blockchain, Ethereum, Ripple, Hyperledger, and Corda, the technology is forging ahead. New Trusted Execution Environments, frameworks, and DLT platforms such as Corda are evolving to address the data privacy and security concerns. And as data privacy and security standards mature, sharing data may even become a prerequisite for obtaining many products and services, including insurance. In the future, connected insurance is likely to become the default; if customers refuse to share data, they may have to pay higher premiums, get some old policy exclusions reimposed, or in extreme cases even be denied a product or service.
Currently, connected insurance is targeting only personal lines, but with the growth of the industrial internet of things, connected insurance may expand to commercial insurance. This expansion may lead to the redefinition of behavioral hazards and preventive measures. In the distant future, when connectedness becomes omnipresent, AI becomes ever more omniscient, and the smart contracts in DLT mature, autonomous insurance products and risk-control measures may emerge. At that stage, moral hazard may become entirely an objective factor in a continuous risk-assessment environment, cease to have any meaningful influence on risk, or be completely eradicated.
But that future, if it comes at all, is a long way off. Until then, the technologies that are changing the world are already making their mark on insurance as we know it—it’s time to think about what’s next.
SRIVATHSAN KARANAI MARGAN works as an insurance domain consultant in Insurance & Healthcare Innovation Lab at Tata Consultancy Services Ltd.
Foundations of Risk Management and Insurance 2nd Edition; Editor: Charles Nyce; American Institute for Chartered Property Casualty Underwriters/Insurance Institute of America; 2006.
Insurance Handbook—A guide to insurance: what it does and how it works; Insurance Information institute; 2010.