The rallying cry has sounded — to “close the protection gap”, the difference between what is paid out by insurance and the total cost of some incident or disaster. Here is an issue that can unite and promote the insurance industry, extending benefits to those in peril by expanding the insurance sector. Having ex-post access to funding after a loss, we know, can bring important benefits.
Yet in reality, there is not just one, but three distinct insurance “protection gaps”, each with separate causes and each requiring different remedies. These protection gaps are so different to one another that we should stop treating them as a single category. Lumping them together can cause confusion.
In this series of four blogs, I will explore each of these three distinct gaps, together with the role of protection gap analytics, and the actions we can plan to address these protection gaps.
The “High-Risk Protection Gap”
In high-risk, high-value regions, the magnitude of potential losses may be so large that insurers and reinsurers do not have sufficient capital to refund a catastrophe. Or maybe they are not allowed, or do not dare, charge what they believe to be the full risk cost and so insurers deliberately restrict the coverage with high deductibles and exclusions. And these restrictions make the policy less appealing to consumers, so a smaller proportion purchase the protection.
Earthquake insurance in California illustrates how the creation of a new entity in response to a recognized protection gap can end up expanding the gap.
In California, private insurers were frustrated by the Insurance Commissioner in their attempt to raise the price of residential earthquake insurance following the unprecedented losses in the 1994 Northridge earthquake. With evidence of new sources of hazard and much higher vulnerabilities than anticipated, the modeled losses had spiked. Any public or private scheme would need to purchase reinsurance and worldwide reinsurance prices were at their highest cost ever after an eight-year run of catastrophes. Yet, the political priority was to maintain the price to consumers.
The solution in 1996 was the establishment of the California Earthquake Authority (CEA), which created the “mini-policy”.The deductible was raised from 10 to 15 percent (the highest for any residential coverage worldwide) and exclusions introduced for some of the previous coverages including “appurtenant structures” outside the main property, which had generated extraordinary levels of claims in the 1994 earthquake. Through the exclusions and raised deductibles, the per property technical risk cost was assessed to have fallen to between 68 to 73 percent of the previous policy, expanding the protection gap by at least 30 percent.
The reduced terms and high deductible also meant that many homeowners dropped the coverage. From 1998 to 2003 the number of policies reduced by more than 20 percent, so the protection gap grew still wider. However, since 2004, as the United States Geographic Survey (USGS) hazard model relaxed its earlier conservatisms, and the price of international reinsurance fell, so the rate charged for the coverage reduced. (Between 1996 to 2013 construction costs grew by 143 percent while the price of earthquake insurance increased by only 29 percent.)
Since 2004, the number of policies has expanded, although not keeping up with the number of homes, so for most of that period, the percentage of homeowners with earthquake cover has continued to fall, reaching 10.8 percent in 2016. Those who buy the coverage tend to be older homeowners, without a mortgage, with more to lose if their house should be destroyed.
Measured in terms of the strength of the CEA’s business model, or the price to consumers, the scheme was a great success. However, it would not score so highly if measured in terms of its societal function to enable rapid recovery after a disastrous earthquake.
For many years it was only earthquakes that provided effective promotion for Californian earthquake insurance. The end-2003 Mw6.6 San Simeon earthquake marked the turnaround in the decline in CEA policy numbers. Distance is a trade-off with size: 8,474 new policies were sold following the 2011 M9 Japan earthquake, while 8,176 householders purchased cover after the 2014 Mw5.1 La Habra earthquake in southern California.
In the Mw6 South Napa earthquake of August 2014, the economic loss was estimated as between US$443 million to US$800 million, of which only one to two percent was covered by insurance — i.e. a gap of 98 to 99 percent, comparable to the poorest countries in the Caribbean.
Chastened by the South Napa experience, the CEA has sponsored building retrofits, more choice around deductibles and offered “discounts” of up to 30 percent on the price of insurance for lower risk locations. As a result, in 2017 the CEA sold 90,707 new policies and an additional 9,000 in January 2018 (more than the average new policies sold for an entire year from 2005 to 2015). However, the expansion of coverage has so far only reduced the protection gap by around one percent.
The protection gap is significantly lower for U.S. flood coverage than for California earthquake, principally because those within the Special Flood Hazard Areas (SFHA) — the FEMA-mapped “100-year flood zone”, are required by the bank to purchase flood insurance when taking out a mortgage. However, the subsequent lapse rate is around 25 percent each year, as no one from the bank checks that insurance is still in place, and the loan has been packaged into a mortgage-backed security. The gap is lower for those at greatest risk, so how you measure the protection gap depends on how you apply the denominator. For those who live in the coastal flood “A” zone, the protection gap is about 30 percent, rising to 50 percent within the 500-year flood zone, and 60 percent across the whole U.S. For the flooding from Hurricane Harvey, around 70 percent of the residential loss was not repaid by insurance.
As with earthquakes, mega-floods prove to be the best marketers of flood insurance. The flooding from Hurricane Katrina helped drive a 17 percent expansion in the number of NFIP flood policies over the following two years, with almost all the increase in the three worst impacted states. Hurricane Harvey’s flooding also caused an expansion in take-up rates, principally in Texas, although not on the scale that followed Katrina. Surprisingly there was no significant uptick in the number of policies following the flooding from Superstorm Sandy.
Outside the U.S., where the peril coverage is a standard addition to fire insurance, as with New Zealand earthquake or French flood, the protection gap is very low. However, where coverage has to be purchased as an extension, even enlightened risk-centered countries, like Switzerland, have wide protection gaps for earthquake.
The insurance sector has a strong vested interest in working to reduce the risk when the level is so high that it challenges the functioning of insurance. We see powerful examples of this in Florida, where new building codes were introduced after 1992 Hurricane Andrew. Following Hurricane Charley in 2004, one study showed for buildings constructed to the new code, a 60 percent reduction in the proportion of homes suffering damage and a 42 percent lowering in the cost of a claim. Another study led by Wharton showed a 72 percent reduction in overall damage costs. The new building code has reduced the overall level of risk, and as a result made insurance more affordable. Both factors in combination are helping to reduce the protection gap.
It is in the long-term interests of society to reduce the high-risk protection gap. There is a cost to the wider economy from a homeowner who cannot repair their property and who defaults on loans, because they did not have insurance coverage.
As widely demonstrated, the best way to close the gap is to provide a standard, technically priced, all perils (fire, wind, flood, earthquake and landslide) homeowners insurance coverage and mandate that this coverage is required to be in place throughout the lifetime of a mortgage.