This has been one of the leading articles on bbc radio news throughout the day.
https://www.bbc.co.uk/news/business-68349396
I’m pretty sure insurers aren’t allowed to ask your race, nor adjust your premium based on it even if they did.
Obviously premiums are based on historical and theoretical risk with I’m assuming vast amounts of data and algorithms behind them, if you live in an area where motor crime or accident claims are higher you expect to have to pay more. It’s always been the way.
If there’s no ethnicity data behind the claims base how can it be racist? Or am I being very naive? Could insurers be loading premiums on certain streets / areas based on data sourced elsewhere which does have ethnicity information?
What would the answer be? Surely you can’t expect premiums to be evened out throughout the country with no location factor? If you’re going to do that do you ignore driver age, experience, even vehicle make / model?
Is car insurance racist?
- integrale_evo
- Posts: 4518
- Joined: Thu Apr 12, 2018 5:58 pm
Is car insurance racist?
Cheers, Harry
Re: Is car insurance racist?
I said it before - my wife's insurance is more just because she's an immigrant. Change her to 'British' and the premiums drop. There's even a specialist insurer for immigrants (Marshmellow?) who claim to offer better rates for them. It's a con.
The artist formerly known as _Who_
Re: Is car insurance racist?
A lot of it is bollocks, I think there should be a far greater right to privacy. Have a British driving licence then leave it at that. What does it matter if you have kids or not, own or rent a house, or what your specific job is.
Yes you can pull statistics out your arse all day and find infinite ways to screw over people with premiums as you can make any statistic fit your own argument. Bit like saying speeding contributes to 27% of fatal accidents, then one can assume that not speeding is more risky as that contributes to 73% of fatalities...
Yes you can pull statistics out your arse all day and find infinite ways to screw over people with premiums as you can make any statistic fit your own argument. Bit like saying speeding contributes to 27% of fatal accidents, then one can assume that not speeding is more risky as that contributes to 73% of fatalities...
How about not having a sig at all?
Re: Is car insurance racist?
I mean, it's shitty journalism TBH. I mentioned this in the other thread - no one is paying £2k/year on a Fiesta. I suspect they've looked at all the quotes, lowest being £400, highest being £5000 (in other words, those mad quotes where they just don't want the business) and said "oh yeah, the average is £2k"
I'd be more interested in the lowest offered quotes for both. I wouldn't be suprised if it was more like £500 for one, £600 for the other - the latter being based on SO many different factors that potentially aren't mentioned; insurance isn't soley based on crime and accidents.
I'd be more interested in the lowest offered quotes for both. I wouldn't be suprised if it was more like £500 for one, £600 for the other - the latter being based on SO many different factors that potentially aren't mentioned; insurance isn't soley based on crime and accidents.
Re: Is car insurance racist?
The statistics are there for a reason, the insurers are not suddenly going to reduce their profits and give higher risk groups a lower premium. If they collect less data it'll just mean they that lower risk drivers will have to pay more and higher risk drivers will have to pay less.Mito Man wrote: ↑Mon Feb 26, 2024 11:07 pm A lot of it is bollocks, I think there should be a far greater right to privacy. Have a British driving licence then leave it at that. What does it matter if you have kids or not, own or rent a house, or what your specific job is.
Yes you can pull statistics out your arse all day and find infinite ways to screw over people with premiums as you can make any statistic fit your own argument. Bit like saying speeding contributes to 27% of fatal accidents, then one can assume that not speeding is more risky as that contributes to 73% of fatalities...
There isn't really any reason for an insurer to screw over a particular group or category, it's just a statistical distribution leading to an mean profit level.
As someone in a low risk category, I'm happy with it the way it is.
Re: Is car insurance racist?
I think they would have to go to some effort to make it racist, so it seems unlikely. The more likely answer to the article is that they take in more and different data than they used to write the article such as crime rate in surrounding areas, or vehicle specific crime. That wouldn't make a good headline though.integrale_evo wrote: ↑Mon Feb 26, 2024 7:08 pm
If there’s no ethnicity data behind the claims base how can it be racist? Or am I being very naive? Could insurers be loading premiums on certain streets / areas based on data sourced elsewhere which does have ethnicity information?
This could turn into a political debate about poverty and fairness, and whether it's fair that the poorest have to pay more, but from an insurance business model perspective I strongly suspect there's nothing to it except statistical analysis of risk.
Re: Is car insurance racist?
Correlation does not equal causation
I imagine a bigger factor than the crime and accident data will be something like number of late or none payments of monthly direct debits leading to cancellation of policies. Basically the more ethnic areas being more deprived and having less consistent cash flow
I imagine a bigger factor than the crime and accident data will be something like number of late or none payments of monthly direct debits leading to cancellation of policies. Basically the more ethnic areas being more deprived and having less consistent cash flow
Re: Is car insurance racist?
I put my Italian ex on my insurance. You can probably guess which way the premium went.
Re: Is car insurance racist?
Did they do comparable test premiums for a white British person in both areas?integrale_evo wrote: ↑Mon Feb 26, 2024 7:08 pm This has been one of the leading articles on bbc radio news throughout the day.
https://www.bbc.co.uk/news/business-68349396
I’m pretty sure insurers aren’t allowed to ask your race, nor adjust your premium based on it even if they did.
Obviously premiums are based on historical and theoretical risk with I’m assuming vast amounts of data and algorithms behind them, if you live in an area where motor crime or accident claims are higher you expect to have to pay more. It’s always been the way.
If there’s no ethnicity data behind the claims base how can it be racist? Or am I being very naive? Could insurers be loading premiums on certain streets / areas based on data sourced elsewhere which does have ethnicity information?
What would the answer be? Surely you can’t expect premiums to be evened out throughout the country with no location factor? If you’re going to do that do you ignore driver age, experience, even vehicle make / model?
Diaspora communities often congregate in poorer parts of cities, because they arrive looking to settle with little money and often can't afford property or rent in rich areas of cities. Car crime is often high in those poorer areas, and anecdotally, I would suggest driving standards are lower too. I live in an area that is only 6% white british/irish and my insurance premiums are sky high, way more than I ever paid when I lived in Camden or Surbiton. If non-white communities are centered around low-income corners of cities with higher crime levels, then of course insurance premiums will be high. Simple.
Re: Is car insurance racist?
I guess with non-UK nationality drivers there's probably a statistical risk that they are not as well trained. I think for some countries you don't have to take a UK test to convert to a UK licence etc. Not that UK drivers are amazing clearly, but there is a certain standard that averaged out over millions of drivers probably contributes to an extra risk factor. The risk might even be due to less available statistics. Eg. the Monaco driving licence might be the hardest to get in the world, but if they've only got a sample of ten people to assess the algorithm will probably still call it a risk increase.