Unfair Discrimination without Disproportionate Impact. As previously defined, unfair discrimination occurs when rating variables that have no relationship to expected loss are used. A hypothetical example could be if an insurer decided to use rating factors that charged those with red cars higher rates, even if the data did not show this. In this case, there would be no disproportionate impact, assuming protected classes do not own a large majority of red cars. Disparate Treatment. Disparate treatment and unfair discrimination are not directly related if we use the Fair Trade Act definition of unfair discrimination. However, in states where rating on protected class is defined to be unfair discrimination, disparate treatment would be a subset of unfair discrimination. In such cases, an insurer would explicitly use protected class to charge higher rates, with the intention of prejudicing against that class. Intentional Proxy Discrimination. If proxy discrimination is defined to require intent, it would be a subset of disparate treatment, whereby an insurer would deliberately substitute a facially neutral variable for protected class for the purpose of discrimination. Redlining is an example of this type of discrimination, given the use of location characteristics as proxies for race and social class. Disproportionate Impact. Disproportionate impact focuses on effect on protected class, even if there is a relationship to expected loss. An example of this is the one mentioned in the AAA study, whereby a rating plan that uses age could disproportionately impact a minority group if those in that minority group tend to have higher risk ages. This disproportionate impact is not necessarily the same as proxy discrimination, since it is likely that even after controlling for minority status, age would have a relationship to expected costs.
Unintentional Proxy Discrimination. If proxy discrimination is defined to be unintentional, the focus is more on disproportionate outcomes and the variables used to substitute for protected class. Several variables are being investigated by regulators to potentially be proxy discrimination and include criminal history for auto insurance rating. In order to prove proxy discrimination, an analysis would have to be performed to understand the extent to which criminal history proxies for minority status, and whether its predictive power would decrease when controlling for protected class. It is important to note once again that terms like “unintentional proxy discrimination” may be subsumed by “disparate impact,” but they are included in this paper to show how various stakeholders use the term differently. Disparate Impact. Disparate impact is unintentional discrimination, where there is disproportionate impact, but also other legal requirements, such as the existence of alternatives. To date, no disparate impact lawsuits against insurance companies have been won. An example of potential disparate impact (although it was not litigated as a lawsuit) is from health care. Optum used an algorithm to identify and allocate additional care to patients with complex healthcare needs. The algorithm was designed to create a risk score for each patient during the enrollment period. Patients above the 97th percentile were automatically enrolled in the program and thus allocated additional care. Upon an independent peer review of the model, researchers found that the model was in fact allocating artificially lower scores to Black patients, even though the model did not use race. The reason behind this was the model’s use of prior healthcare costs as an input. Black patients typically spend less than white patients on health care, which artificially allocated better health to Black patients.18 Unfair Discrimination and Disproportionate Impact. In this case, an insurer would use a variable that both has no relationship to expected loss, but also has an outsized effect on protected classes. An example of this could be the same red car case above, but where protected classes also owned almost all the red cars. In this case, higher rates would create a disproportionate effect on protected classes, while also having no relationship to expected loss.
Insurers are bracing for a hit of between $28 billion and $47 billion from Hurricane Ian, in what could be the costliest Florida storm since Hurricane Andrew in 1992, according to U.S. property data and analytics company CoreLogic.
Wind losses for residential and commercial properties in Florida are expected to be between $22 billion and $32 billion, while insured storm surge losses are expected to be an additional $6 billion to $15 billion, according to CoreLogic.
“This is the costliest Florida storm since Hurricane Andrew made landfall in 1992 and a record number of homes and properties were lost,” said Tom Larsen, associate vice president, hazard & risk management, CoreLogic.
The private market for flood insurance in the United States measures approximately $300 million in annual premium. This is less than 10 percent of the $3.7 billion in flood insurance premium written by the federal government’s National Flood Insurance Program (NFIP). Private insurers offering flood insurance are not operating on the same playing field because many NFIP policies are subsidized and underpriced. The creativity of private insurers, guided by the dynamics of a free and competitive market, will eventually drive out inefficiency and false price signals, and make available to homeowners and businesses the flood insurance they need at the right cost.
We invite you to an online discussion examining the obstacles and opportunities for private insurers featuring flood insurance entrepreneur Trevor Burgess, and R Street’s Jerry Theodorou and Caroline Melear.
Author(s): Jerry Theodorou, Trevor Burgess, Caroline Melear
Mr. Toomey asked Jerry Theodorou of the R Street Institute, a conservative-leaning Washington-based think tank, how seriously Congress should look at paying repetitive loss claims. “Indeed, this is a very serious problem,” Mr. Theodorou said. “The numbers speak for themselves, to have such a small percentage of policyholders accounting for close to 40% of the claims dollars paid.”
The flood insurance program — which is the main provider of flood coverage in the U.S. and has issued more than five million policies — has paid out more money to property owners and other expenses than it has collected in premiums from policyholders since Congress created it in 1968. It collects about $4.6 billion in annual revenue from policyholders in premiums, fees and other charges, according to the Congressional Research Service.
Flooding ranks as the country’s most common natural disaster. Scientists predict floods will happen more frequently in neighborhoods that face new risk from rising sea levels and extreme rainstorms due to climate change.
Regarding the second objective, there is no equitable sharing of costs between the public and private sectors. The private sector is only peripherally involved in bearing flood risk. The involvement of the private insurance sector is restricted to administration of the program, for which insurers are remunerated by the NFIP. The participation of private insurers in flood insurance as a risk-bearer is de minimis, writing less than a tenth the premium collected by the NFIP.
Instead of attaining the overarching goal of reducing economic losses caused by flooding, flood- related economic losses have increased. In the past decade, U.S. economic losses caused by flooding were $943 billion, close to five times more than the $211 billion cumulative flood-related losses in the prior decade. In this testimony, we highlight five issues standing in the way of the NFIP falling short of achieving its mission, and propose solutions to remedy those problem areas.
The American Academy of Actuaries presents this summary of select significant regulatory and legislative developments in 2021 at the state, federal, and international levels of interest to the U.S. actuarial profession as a service to its members.
The Academy focused on key policy debates in 2021 regarding pensions and retirement, health, life, and property and casualty insurance, and risk management and financial reporting.
Responding to the COVID-19 pandemic, addressing ever-changing cyber risk concerns, and analyzing the implications and actuarial impacts of data science modeling continued to be a focus in 2021.
Practice councils monitored and responded to numerous legislative developments at the state, federal, and international level. The Academy also increased its focus on the varied impacts of climate risk and public policy initiatives related to racial equity and unfair discrimination in 2021.
The Academy continues to track the progress of legislative and regulatory developments on actuarially relevant issues that have carried over into the 2022 calendar year.
Insurance giants Chubb, Liberty Mutual, and AIG are three of the biggest insurers of fossil fuel infrastructure around the world. But thecompanieshave just announced plans to scale back their homeowner coverage in California, where they insist future climate-related losses will likely prevent them from turning a profit.
The coverage withdrawals may soon ignite a big money battle in the state’s legislature, pitting insurance giants against lawmakers trying to preserve coverage for their constituents. Meanwhile, climate campaigners are decrying what they say is a fundamental hypocrisy.
Last year, Chubb’s chairman and CEO Evan Greenberg said the company was reducing its coverage in parts of the state that were “both highly exposed, and even moderately exposed, to wildfire” because it was unable to obtain an “adequate price for the risk, and not by a small amount” due to both the costs of wildfires and California’s regulatory climate.
A main solution proposed by industry is that they be allowed to use “catastrophic modeling,” a method where rates are set based on predictions of future losses, rather than recorded past losses, as is currently the case. All other states allow the use of this technique in at least some cases.
Nearly a quarter of U.S. critical infrastructure—utilities, airports, police stations and more—is at risk of being inundated by flooding, according to a new report by First Street Foundation, a Brooklyn nonprofit dedicated to making climate risk more visible to the public.
Roughly 14% of Americans’ properties face direct risk from major storms, but the study shows danger extends far from those property lines.
The authors say the report provides the first holistic understanding of flood risk beyond individual property level. In addition to critical infrastructure, the report assesses commercial buildings, millions of miles of roads and socially important institutions such as schools and museums.
“Even if your home is far from the risk of flooding or forest fires, you may not so easily escape the systemic impacts from vulnerable critical infrastructure that sometimes extends hundreds of miles,” said Jesse Keenan, a climate-change and real-estate expert at Tulane University in New Orleans.
Author(s): Leslie Kaufman, Rachael Dottle, Mira Rojanasakul
The long-term security of coastal regions depends not simply on climate, oceans and geography, but on multiple local factors, from the politics of foreign aid and investor confidence, to the quality of resilience-oriented designs and ‘managed retreat’.
Take some examples. In 2017, the drought in Cape Town and lack of resilient water infrastructure led to a downgrade by Moody’s. Wildfires in the Trinity Public Utilities District in California led to similar downgrades in 2019. Moody’s have developed a ‘heat map’3 that shows the credit exposure to environmental risk across sectors representing US$74.6 trillion in debt. In the short term, the unregulated utilities and power companies are exposed to ‘elevated risk’. The risks to automobile manufacturers, oil and gas independents and transport companies are growing. Blackrock’s report from April 2019, focused primarily on physical climate risk, showed that securities backed by commercial real estate mortgages could be confronted with losses of up to 3.8 per cent due to storm and flood related cash flow shortages.4 Climate change has already reduced local GDP, with Miami top of the list. The report was amongst the first to link high-level climate risk to location analysis of assets such as plants, property and equipment.
In other words, adaptation and resilience options are also uniquely local. The outcomes hinge on mapping long-term interdependencies to predict physical world changes and explore how core economic and social systems transition to a sustainable world.
Science has provided America with a decent idea of which areas of our country will be most devastated by climate change, and which areas will be most insulated from the worst effects. Unfortunately, it seems that population flows are going in the wrong direction — today’s new Census data shows a nation moving out of the safer areas and into some of the most dangerous places of all.
Some of the examples are genuinely mind-boggling. For instance, upstate New York is considered one of the country’s most insulated regions in the climate crisis — and yet almost all of upstate New York saw population either nearly flat or declining. At the same time, there were big population increases in and around the Texas gulf coast, which is threatened by extreme heat and coastal flooding.
Similarly, the city of Philadelphia is comparatively well situated in the climate crisis — but it saw only modest population growth of 5 percent. It was surpassed on the list of biggest cities by Phoenix, which saw an 11 percent population growth, despite that city facing some of the worst forms of extreme heat and drought in the entire country.
Theoretically, wealthier people should buy less insurance, and should self-insure through saving instead, as insurance entails monitoring costs. Here, we use administrative data for 63,000 individuals and, contrary to theory, find that the wealthier have better life and property insurance coverage. Wealth-related differences in background risk, legal risk, liquidity constraints, financial literacy, and pricing explain only a small fraction of the positive wealth-insurance correlation. This puzzling correlation persists in individual fixed-effects models estimated using 2,500,000 person-month observations. The fact that the less wealthy have less coverage, though intuitively they benefit more from insurance, might increase financial health disparities among households.