Heikki Vesanto, Manager, GIS data science, LexisNexis Risk Solutions, Insurance, UK and Ireland
This is where a score-based approach to risk assessment can help smooth out crime trend anomalies so that insurance providers can price home and commercial property insurance down to a postcode or commercial address as swiftly, fairly and competitively as possible.
Pandemic claims trends - here to stay or gone tomorrow?
The past two ‘pandemic’ years saw reductions in many types of crimes including shoplifting, robbery, and burglary . Whether this trend continues for the long-term remains to be seen but for now we have not seen a return to pre-pandemic levels for many crime types, although there are regional differences. There seems to be a clear correlation between home occupancy, the work from home trend, and reported burglaries that looks set to be a factor for future home insurance pricing - indeed, the proportion of people hybrid working has risen in 2022 .
Overcall crime rates down but arson spikes
Perhaps not surprisingly, lockdowns had a significant impact on overall criminal behaviour , not just burglaries but during the third lockdown in 2021, criminal damage and arson rose perhaps an outcome of the cumulative financial distress of the whole time period.
Arson is a growing concern for property insurance providers and accounts for 50% of all fires attended to by the fire services. The National Fire Chiefs Council (NFCC) puts the reported cost of fire insurance claims at £1.2bn and it’s a category of losses that’s been on an upward trend.
Bearing in mind criminal damage and arson spiked in Lockdown three, possibly due to financial stress, now more than ever, insurance providers need to understand fire and arson risks to help ensure they can price accurately and fairly.
Arson and Burglary Scores
Arson risk can now be assessed down to the level of the outline of a building, looking at factors such as any at-risk properties nearby considering prior loss and events, fire service response times and distance to the nearest fire station for greater underwriting precision.
Crime modelling for a post-pandemic world
The crime model on which these scores are based utilises over 65 million crime events going back to 2010 and is refreshed every six months, to rank postcodes for crime risk. Crimes are categorised and located spatially while advanced spatial modelling ties the crime event data to postcodes. This spatially smooths the crime data to provide a better indication of the crime level in an area.
Population data is also used as a control, to make sure high-density areas do not unduly score highly.
Access to the most up to date intelligence on local crime rates alongside data on perils and granular property characteristics data down to the individual address can all build a more comprehensive picture of property risk, at the point of quote, to help support pricing accuracy at new business and renewal. Furthermore, burglary and arson scores have shown in tests to be a good indicator of loss propensity, with each increase in crime score corresponding with an increased loss relativity. Advances in data enrichment technology mean that all this data can be pulled into quoting systems or mapping solutions from one platform at sub-second speed, for a total view of risk around the property, the person or business and the place.
While the pandemic overturned a lot of the historic assumptions around crime and the drivers for crime related losses, it can be fair to consider that most fluctuations in recorded crime over the past two years can be attributed to the introduction and subsequent easing of national lockdown restrictions alongside more normal seasonal variations for some crime types. Crime scores smooth out these bumps to help ensure the fairest assessment of risk for pricing.
|
