00:00:00
There are quite a few discussions around
insurance these days.
00:00:03
One that's been talked about but maybe not
enough and is getting more and more
00:00:07
focused now is how credit
scores can have an impact
00:00:11
on how much you pay
for homeowners insurance.
00:00:14
Pleasure to be joined
by Ben Keys, Professor
00:00:16
of Real Estate here at
the Wharton School.
00:00:18
Ben, great to catch up again.
How are you, sir?
00:00:20
I'm doing well.
Thanks so much for having me, Dan.
00:00:23
Yes, absolutely.
All right.
00:00:24
So the bigger question I think I have
right at the top is do people realize the
00:00:30
level of impact that a credit score can
play on your homeowners insurance rates?
00:00:36
Yeah, I don't think they do.
00:00:37
I was certainly surprised by our findings
in this study looking at households with a
00:00:43
very rich data set to see
just how much their credit
00:00:46
score impacts the cost
of homeowners insurance.
00:00:49
And it turns out it's quite a lot.
00:00:51
We've had many discussions over the last
few years about climate risk and some of
00:00:56
the rising costs there in terms of where
you live.
00:00:59
But they're not just pricing on place.
00:01:01
They're also pricing on people.
00:01:03
Right.
00:01:04
And so there's a couple of components to
discuss here.
00:01:07
Speaking about the people, is there a
significant difference when you're talking
00:01:13
about credit score with lower and higher
income families?
00:01:18
Yeah, I mean, so certainly there's a
strong relationship that people with
00:01:21
higher incomes have
higher credit scores, but
00:01:24
it's certainly not a
one-to-one relationship.
00:01:27
One of the things that we're struck by in
this analysis is thinking about what
00:01:31
credit scores are proxying for when
insurers use this in their pricing.
00:01:35
And so they're not allowed to use credit
scores in all states.
00:01:39
And we'll talk more about some of the
variation across the country in this.
00:01:44
But when you're looking at this,
one of the challenges is, well,
00:01:47
credit score is correlated with a lot of
other things and a lot of other attributes.
00:01:51
And so from a research standpoint,
we wanted to try to find a setting where
00:01:55
we could isolate just a change in the
ability of insurers to use credit scores,
00:02:01
rather than all of the other things that
credit score might be correlated with.
00:02:05
Now, the other component which you and I
have talked about is climate risk.
00:02:09
And your research talks about the fact
that credit score, in many cases,
00:02:14
is an even bigger burden at times than the
climate risk component.
00:02:19
Yeah, we were very struck by this,
that if you look across much of the
00:02:22
country— so not in the most risky parts of
the country, but if you look at the rest
00:02:27
of the country, you know, the vast
majority of the distribution of households
00:02:30
in the US, your credit score is actually a
bigger determinant than your climate risk.
00:02:36
If you were to just improve your credit
score a little bit, that's going to have a
00:02:40
bigger impact on your
cost of insurance relative to
00:02:44
just doing a small reduction
in your climate exposure.
00:02:48
And so if you could think about the kind
of the gradient between your expected
00:02:51
losses or the gradient
between your credit score, it's
00:02:54
that credit score relationship
that's actually steeper.
00:02:57
And that might undermine sort of some of
the incentives to reduce your risk or to
00:03:03
mitigate some of the hazards when you
might be better off paying down some of
00:03:07
your credit card debt and improving your
credit score.
00:03:09
That might be a quicker way to lower your
insurance bills.
00:03:13
Do we know how much we're talking about
here differential, whether it be
00:03:17
percentage or actually dollars wise,
of the type of difference we're seeing
00:03:22
with people that are paying
more on their homeowners
00:03:25
insurance than people
that are paying less?
00:03:28
Yeah, so in our study we find that, all
else equal, households in the top quartile
00:03:33
of credit scores pay
about $550 a year less
00:03:36
for insurance than
those in the bottom 20%.
00:03:40
And that's the average.
00:03:41
That varies quite a bit across the
country.
00:03:44
And it's quite a bit larger difference in
the highest risk areas.
00:03:48
So there's kind of a double whammy for
living in risky areas, that if you have a
00:03:53
riskier location and also
a lower credit score, you're
00:03:57
going to pay substantially
more for homeowners insurance.
00:04:00
Now, I'm sure there are going to be people
out there that are going to say that
00:04:04
laying this out the way you have,
it's a bit of an unfair process for those
00:04:08
at the lower end of the spectrum.
00:04:10
What do you say to them
about how to kind of deal
00:04:14
with this and mitigate it
the best way that they can?
00:04:17
Well, this is the tension in setting
risk-based pricing.
00:04:21
And three states have banned the use of
credit scores in setting homeowners
00:04:26
insurance rates, California, Maryland and
Massachusetts.
00:04:29
And the view in those states is that it
isn't a fair practice to use credit
00:04:35
information to set your homeowners
insurance.
00:04:38
In the other 47 states in
the country, you are allowed
00:04:42
to use credit scores,
although to varying extents.
00:04:45
And I think what that suggests is that
there's a variety of opinions on whether
00:04:50
this is fair from a sort of policy
standpoint.
00:04:54
I think actuarially, the insurance industry
says, "Well, there's predictive power in
00:04:58
the credit score.
00:04:59
It's proxying for a bunch of things.
00:05:02
And if we try to unpack what some of those
things are that it's proxying for,
00:05:07
you know, one might be that people with
lower credit scores live in riskier areas.
00:05:12
That's something that
we can account for very
00:05:14
well in our rich data from
McDash and CoreLogic.
00:05:19
Some of it might be that they live in
older homes.
00:05:23
And so there may be a difference there in
terms of the potential risks.
00:05:27
That's another thing that we can account
for.
00:05:28
But I think there's a couple of kind of
remaining concerns that are out there.
00:05:32
One is related to household liquidity,
just the likelihood of filing a claim
00:05:38
after an event rather than using your
savings to cover it and sort of
00:05:42
recognizing that multiple
filings has a negative
00:05:45
impact on the cost of
insurance going forward.
00:05:48
So that dynamic story. Or it might be kind
of a proxy for conscientiousness,
00:05:53
that the people who, you know,
miss a credit card payment are the same
00:05:56
people who might not be maintaining their
homes at the same quality level.
00:06:00
You know, the insurance industry doesn't
necessarily unpack those different drivers.
00:06:04
They don't have the data to do so.
00:06:07
And so they use this catchall of a credit
score as a proxy for these things.
00:06:13
And you can see in their models,
they argue that it's a strong predictor of
00:06:17
ultimately the losses on a given property.
00:06:21
So that's the trade off is, you know,
this idea of fairness is from actuarial
00:06:25
standpoint, is this seen as fair, versus
politically, or socially,
00:06:30
do we think that this is fair?
00:06:32
And, you know, when you see examples of
next door neighbors paying, you know,
00:06:35
2x what the other one is
paying, you start to think,
00:06:38
well, that doesn't— doesn't
feel like a very fair system.
00:06:41
What do you think, then, are some of the
solutions?
00:06:44
And is there a potential federal solution?
00:06:47
Because obviously, most of this work is
being done at the state level.
00:06:52
You know, for the federal
government to try and help
00:06:55
address some of these
challenges in the insurance market.
00:06:58
Yeah, the tension is that the states
really, you know, are the ones who
00:07:02
regulate homeowners insurance,
and the federal government
00:07:05
only plays a very small
role, if at all at the moment.
00:07:09
And so this is a state-by-state process to
determine, you know, what is the— what is
00:07:14
the ability of the insurers to use
additional information?
00:07:19
And what is this actually proxying for?
00:07:21
What is it capturing?
00:07:22
I think we need better data from the
industry to connect this to claims data
00:07:26
and to understand what some of the
underlying drivers are of the pricing
00:07:31
relationship that they see that they're
then passing on to the homeowners in
00:07:36
the form of higher premiums.
00:07:38
What about the federal reinsurance market?
00:07:41
So the federal market doesn't—
you know, there isn't really a federal
00:07:44
reinsurance market when it comes to
homeowners insurance.
00:07:50
There's a federal flood
program, and they don't
00:07:54
price credit scores in
this same way.
00:07:57
So you don't see the same relationship in
the national flood insurance program.
00:08:01
But I think, you know, in terms of
addressing sort of credit-score-related
00:08:05
issues, it really does become a state by
state story.
00:08:09
And in our paper, we dig into a policy
change that happened in Washington state
00:08:13
where they essentially flipped the switch
and said you can't use credit scores and
00:08:18
then flipped the switch back after intense
lobbying from the industry.
00:08:21
And there, we could kind of hold a lot of
the other characteristics constant.
00:08:25
These were the same
homeowners in the same houses
00:08:27
with the same insurance
carriers in most cases.
00:08:30
And we can see exactly
what the effects of taking credit
00:08:33
scores out of the pricing model
did to the cost of insurance.
00:08:38
But the challenge for the consumer
continues to be one where there's not a
00:08:42
lot, it seems like, that they can do on
their end to be able to have a significant
00:08:48
impact to make this system work better
down the road, it feels like.
00:08:53
These are challenging times in the
homeowners insurance market.
00:08:57
We are seeing some softening in a few of
the most expensive spots in the country.
00:09:02
But in general, insurance prices are up
quite a bit.
00:09:06
The McDash data suggests
that they're up almost 7%
00:09:09
over the last year, which
is far faster than inflation.
00:09:12
And that's after multiple years of double
digit growth.
00:09:16
And ultimately, what our study is pointing
to is that this insurance burden falls
00:09:20
most heavily on the lowest credit score
households.
00:09:23
And so those are the
households that have the least
00:09:25
cash on hand and that
need insurance the most.
00:09:27
They're the ones who are
going to benefit the most from
00:09:29
having this protection in
place in the event of a disaster.
00:09:33
And they're the ones who can afford it the
least.
00:09:37
And so that's really the tension in this
market right now.
00:09:40
Ben, great to talk to you again,
as always.
00:09:41
Thanks very much.
00:09:43
Yeah, thanks so much for having me,
Dan.
00:09:44
Thank you.
00:09:45
Ben Keys, Professor of Real Estate here at
the Wharton School.
