I own neither the study nor the data it was based upon, but the key is to compare loss ratios, not claim counts. Prior to the application of credit, but after the application of adequate territory (and all other) relativities, those with higher credit scores have consistently lower loss ratios.
That said, given the correlation between low credit score and high territory relativity, it is important that a multi-variate analysis is used to avoid the double-counting you mention. Any actuary worth his salt knows that and would have taken this step to ensure that when both territory relativity and credit score factor are applied, any correlation between the two is considered and there is no unfair discrimination. Anything less would expose the actuary to a disciplinary hearing for violating professional standards.
I always like how these uninformed politicians jump on an issue, they clearly do not understand, in order to further their political careers at the cost of the very constituents they purport to represent. I guess all the companies that provide insurance jobs in Michigan are less important to consider than cheaper insurance rates. Oh, and that would be cheaper rates for lifetime medical benefits. The most liberal program in the country. Discontinue PIP and you can cut the rates by 50% rather than 20%.
Clarke takes over where Granholm leaves off; the only industry left that she hasn’t ruined is the insurance industry but between the two of them, they will take care of that. Detroit gets lower insurance rates, I want lower food prices, gas, newspaper etc. They never think about us northerners and the higher costs we have to pay for everything else but insurance.
I still have to wonder why the credit/insurance scoring issue has been blown so far out of proportion. Banks charge intrest base on risk. Banks have solid data that shows people with certain characteristics on a credit report represent a higher risk and therefore deserve a higher intrest rate. This data is not scientific, merely statistical.
The insurance industry has learned, using similar statistical formulas, that characteristics in a credit file are an indicator for future insurance loss. Thus assuming the applicant is a higher risk.. Exactly the same as banks.
I’ve never heard anyone accusing the banks of “high-tech” redlining.
Why???? Because they’ve been doing it for 100 years. The only reason the Insurance Industry has gotten a bad rap is because it’s something new.
PIP plays a lot in the role of Michigan auto insurance. It’s not the only factor, but it plays a big part. In Florida we only have $10,000 in PIP coverage available and many areas here are nearly as expensive as the urban areas of Metro Detroit. The whole no-fault system is flawed in my opinion.
Mr. Risk, you’re being too logical. You can’t argue with these people that think credit is “unfair” or not relevant. It’s a complete emotional issue with this group. While I’ve reviewed plenty of data that shows there is a correlation between a person’s financial habits/responsibility/score and their loss experience, NO ONE that is opposed to credit scoring has been able to provide data that shows it’s not relevant. Their claim is that it’s
just “unfair”. Now that’s a solid argument (from an 8 yr old).
I wrote my thesis on the practice of credit scoring. I can show you studies that place risks on a map. In the areas with the highest levels of “poor” credit risks you’ll find that the insurance companies have already factored in higher comp, collision, and PIP rates due to loss ratios. It’s an attempt to double dip. We’ll raise your rates due to territory (and rightfully so) and we’ll raise your rates because you have limited or bad credit. You’ll find that territory, driving record, and claims history are all very powerful in determining risk. A large proportion of the higher credit risk comes from the higher territory risk. The only reason there is a correlation at all is because the studies that I’m sure you’ve seen don’t factor in territory.
This is simply not true. I have personally been involved in a study of this for a former employer. Credit is a significant predictor of losses even after considering every other rating variable. If that weren’t the case, nobody would be using it.
I don’t know why everyone thinks the insurance industry is populated by criminals, but I’ve run across very few dishonest people in my 20 years in the industry.
Actuary, I’d really like to see your study because I’d like to believe you’re telling the truth. Unfortunately I’ve seen studies showing people residing in the 3XXX5 zip code file more claims than those residing in 3XXX6. People in 3XXX5 already have rates 15% higher than those in 3XXX6 because of loss history. It just so happens that those in 3XXX5 also have limited credit or bad credit. Is it because of their credit that they’ve filed the claims or is it because of the territory? Since I am not an actuary I don’t know. I do know that now those in 3XXX5 are hit twice because of their insurance risk score and their territory while most of those in 3XXX6 enjoy the benefit of an additional discount because of credit.
It’s a tricky issue that won’t be resolved overnight. Credit as a variable is far too open to interpretation for my liking. I think using credit as a very small variable in the initial underwriting process and then using internal data only (payment and claims history along with driving record) is the most fair way to go about rating if you’re going to use credit. Unfortunately I feel that far too many companies use outside credit and weigh it too heavily. Only one company that I know of uses credit on the way in and internal data only after 2 years.
Putting data and analytics aside, I can’t help but wonder why companies would spend so much money on the credit technology (it cost money to order these reports) and take on the political fights to defend the use of credit if it’s not a valid loss predictor. Common sense says it must have some validity or companies would toss it out, advertise like hell that they don’t use it (while others do), and save money on credit orders in the process.
In states where credit is allowed to be used, there aren’t many companies that don’t use it and the one’s that don’t, do not have attractive rates or a large market presence. That should be some indicator that it’s not all bad for consumers.
I own neither the study nor the data it was based upon, but the key is to compare loss ratios, not claim counts. Prior to the application of credit, but after the application of adequate territory (and all other) relativities, those with higher credit scores have consistently lower loss ratios.
That said, given the correlation between low credit score and high territory relativity, it is important that a multi-variate analysis is used to avoid the double-counting you mention. Any actuary worth his salt knows that and would have taken this step to ensure that when both territory relativity and credit score factor are applied, any correlation between the two is considered and there is no unfair discrimination. Anything less would expose the actuary to a disciplinary hearing for violating professional standards.
It sounds like Michigan wants to become another NJ or MA…or the auto equivalent of FL’s property market.
I always like how these uninformed politicians jump on an issue, they clearly do not understand, in order to further their political careers at the cost of the very constituents they purport to represent. I guess all the companies that provide insurance jobs in Michigan are less important to consider than cheaper insurance rates. Oh, and that would be cheaper rates for lifetime medical benefits. The most liberal program in the country. Discontinue PIP and you can cut the rates by 50% rather than 20%.
Hopefully it’s a start to eliminating more high-tech redlining (credit scoring). Good luck Hansen Clarke.
Hey Eliminate Credit Scoring – – –
You should do your homework before making such a ridiculous comment!
Clarke takes over where Granholm leaves off; the only industry left that she hasn’t ruined is the insurance industry but between the two of them, they will take care of that. Detroit gets lower insurance rates, I want lower food prices, gas, newspaper etc. They never think about us northerners and the higher costs we have to pay for everything else but insurance.
Remember this in November..”If they are in, they are out”. Certainly applies with Clarke.
I still have to wonder why the credit/insurance scoring issue has been blown so far out of proportion. Banks charge intrest base on risk. Banks have solid data that shows people with certain characteristics on a credit report represent a higher risk and therefore deserve a higher intrest rate. This data is not scientific, merely statistical.
The insurance industry has learned, using similar statistical formulas, that characteristics in a credit file are an indicator for future insurance loss. Thus assuming the applicant is a higher risk.. Exactly the same as banks.
I’ve never heard anyone accusing the banks of “high-tech” redlining.
Why???? Because they’ve been doing it for 100 years. The only reason the Insurance Industry has gotten a bad rap is because it’s something new.
PIP plays a lot in the role of Michigan auto insurance. It’s not the only factor, but it plays a big part. In Florida we only have $10,000 in PIP coverage available and many areas here are nearly as expensive as the urban areas of Metro Detroit. The whole no-fault system is flawed in my opinion.
Love it!!
Mr. Risk, you’re being too logical. You can’t argue with these people that think credit is “unfair” or not relevant. It’s a complete emotional issue with this group. While I’ve reviewed plenty of data that shows there is a correlation between a person’s financial habits/responsibility/score and their loss experience, NO ONE that is opposed to credit scoring has been able to provide data that shows it’s not relevant. Their claim is that it’s
just “unfair”. Now that’s a solid argument (from an 8 yr old).
I wrote my thesis on the practice of credit scoring. I can show you studies that place risks on a map. In the areas with the highest levels of “poor” credit risks you’ll find that the insurance companies have already factored in higher comp, collision, and PIP rates due to loss ratios. It’s an attempt to double dip. We’ll raise your rates due to territory (and rightfully so) and we’ll raise your rates because you have limited or bad credit. You’ll find that territory, driving record, and claims history are all very powerful in determining risk. A large proportion of the higher credit risk comes from the higher territory risk. The only reason there is a correlation at all is because the studies that I’m sure you’ve seen don’t factor in territory.
This is simply not true. I have personally been involved in a study of this for a former employer. Credit is a significant predictor of losses even after considering every other rating variable. If that weren’t the case, nobody would be using it.
I don’t know why everyone thinks the insurance industry is populated by criminals, but I’ve run across very few dishonest people in my 20 years in the industry.
Actuary, I’d really like to see your study because I’d like to believe you’re telling the truth. Unfortunately I’ve seen studies showing people residing in the 3XXX5 zip code file more claims than those residing in 3XXX6. People in 3XXX5 already have rates 15% higher than those in 3XXX6 because of loss history. It just so happens that those in 3XXX5 also have limited credit or bad credit. Is it because of their credit that they’ve filed the claims or is it because of the territory? Since I am not an actuary I don’t know. I do know that now those in 3XXX5 are hit twice because of their insurance risk score and their territory while most of those in 3XXX6 enjoy the benefit of an additional discount because of credit.
It’s a tricky issue that won’t be resolved overnight. Credit as a variable is far too open to interpretation for my liking. I think using credit as a very small variable in the initial underwriting process and then using internal data only (payment and claims history along with driving record) is the most fair way to go about rating if you’re going to use credit. Unfortunately I feel that far too many companies use outside credit and weigh it too heavily. Only one company that I know of uses credit on the way in and internal data only after 2 years.
Putting data and analytics aside, I can’t help but wonder why companies would spend so much money on the credit technology (it cost money to order these reports) and take on the political fights to defend the use of credit if it’s not a valid loss predictor. Common sense says it must have some validity or companies would toss it out, advertise like hell that they don’t use it (while others do), and save money on credit orders in the process.
In states where credit is allowed to be used, there aren’t many companies that don’t use it and the one’s that don’t, do not have attractive rates or a large market presence. That should be some indicator that it’s not all bad for consumers.