AALS Section on Insurance Law: Insurance and Consumer Protection

Moderator:  Joshua C. Teitelbaum, Georgetown University Law Center

Speakers:  Orly Lobel, University of San Diego School of Law,

Kyle D. Logue, University of Michigan Law School

Daniel Schwarcz, University of Minnesota Law School

Lauren E. Willis, Loyola Law School

Lobel (paper with On Amir), Risk Management for the Future: Age, Risk, and Choice Architecture

If we’ve shifted to individual choice, consumer autonomy and private decisionmaking, how can policy direct sound choices.  A lot of insurance/financial planning decisionmaking is suboptimal. People don’t save optimally and rigorously for the future.  One recent study: American households have ¼ of what they need to maintain a good financial level for older age.

Dual system model: system 1 is fast, intuitive, affective, automatic, reflexive; system 2: slower, cognitive, reflective.  Executive resource depletion: prior unrelated tasks exerting executive control may hinder subsequent activities, hindering system 2’s ability to override. Risk tolerance and risk aversion are different in the systems: hindering system 2 changes risk-taking.  We don’t just assume older people are more risk-averse: there are different states in each person, and the decisionmaking environment interacts with that.

Experiments: hinder system 2 with previous task before focal task.  Phonetic matching v. Stroop task (make decisions about color of color words written in different color), the latter designed to deplete executive function.  Then asked about preferences about retirement savings.  Younger people are less likely to opt into savings.  When we deplete both groups, though, the younger doesn’t change significantly but there’s a significant difference in the depleted older groups; older depleted subjects are less likely to opt in.

Annuities v. lump sums: younger people opted for the annuity versus the lump sum more than older people across all conditions.  Possible explanation: The closer you are to getting the pile of money, the more tempted you are.  Depleted: both groups are more likely to want the lump sum.  Not enough literature about what to do with the money once people retire—annuities v. lump sums.

Stocks v. bonds.  Similar results.  It may be that the risk-aversion of older people is System 2 at work.  Older groups are more likely to fall prey to gambling, addiction, etc.; when decisionmaking environment is a sound one, helps people use System 2, older people do better/override automatic/natural tendencies.

What happens when you offer three options instead of two?  Depleted younger people even more frequently go for the cheaper option and middle option over the more expensive/safer option.  But the depletion effect was less pronounced for older groups (in fact, it looked to me like the depleted older group increased its choice of the most expensive option at the expense of the least expensive).  When a decision is important enough, depletion is harder.  The middle effect grows in relation to the cheap one.

When cognitive resources are available, older participants opt for more prudent savings choices. Much less so when depleted—which could apply to sickness, fatigue, distraction, or after making other arduous choices.  In some instances, depletion affects older people more than younger.  Policy should be tailored: most important decisions are up front and there are sticky defaults/path dependence.

Kyle Logue: Outsourcing Regulation: How Insurance Reduces Moral Hazard (with Omri Ben-Shahar)

How insurance companies work to reduce risk.  Insurance is privatized regulation, making insurers regulators.  First party or third party insurance is an alternative to, and can be superior to, agency-based regulation.  Contrast with conventional wisdom: insurance is only risk spreading.  Our research: aspects of private insurance market that give it a comparative advantage as regulators.  Discussions with insurers: not clear how much they actually use the research that they show off to academics; they are hesitant to share data.

Why do insurers want to regulate risk?  Don’t they profit more when risk is high?  Yes, to a point, though they have some incentive to reduce risk in the long run, though not to zero—there have been debates in the industry over this.  To a point, they just price the risk, but they also try to get policyholders to do less risky things and they lobby for certain regulations that reduce risk.  Some of the reason is competition; maximizing size of market; avoiding regulatory overreaction if rates get too high.

Insurers have access to information, often better than regulatory agencies, through the underwriting process; etc. Insurers can engage in ex ante regulation through risk-rated premiums (privatized Pigouvian tax), discounts for good behavior; deductions and copayments; refusal to insure/gatekeeping; coaching safer conduct; lobbying gov’t for risk-reducing regulation. Ex post regulation: claims management/enforcing exclusions; requiring insureds to mitigate losses.

Auto insurance: you’ve seen the Progressive ads etc.  Trying to get data from consulting firm for big auto companies about ways they can be incentivized to reduce risks.  Some attempts are just attempts to find the low-risk people; others might be lowering risks.

Speculating about new domains for insurance: first-party insurance replacing contract litigation with consumer contract insurance?  There’s not much threat of regulation, and compelled mandatory arbitration makes it hard to sue.  If we aren’t going to dramatically ramp up the budget for the FTC/CPSC, why not insure people against small risks. eBay Motors, SquareTrade, VRBO (vacation rental by owner).  This might expand.

Limits: when criminal sanctions are needed; when risks are uninsurable; when risks are insured primarily through group policies; when there are information externalities (hard to patent insurance techniques); inter-temporal problems like cliamate change; when insurance is considered an entitlement (health insurance, which we have left aside entirely?).  Also, how do we protect consumers from insurers?

Schwarcz: Transparently Opaque: Understanding the Lack of Transparency in State Insurance Consumer Protection

If you look at consumer protection in insurance, it’s vastly different than in other domains in ways that raise substantial normative problems.

Transparency can be achieved by summary disclosure, dominant in banking & credit (APRs), important in securities and health insurance: core rationale—empower consumers.  Full disclosure—dominant in securities; important in banking and health insurance.  Core rationale: empower sophisticated intermediaries.  Consistent with and complementary to more aggressive forms of regulation; disclosure doesn’t have to substitute for other regulations.

But it’s impossible to find out which insurer pays more claims or pays faster.  Regulators have these data but don’t make them publicly available.  Effective regulatory solution would combine transparency with other regulation.  Make sure what’s actually in the contract is the coverage consumers reasonably expect.  When you buy an insurance policy, you don’t get the summary policy until 2 weeks later. There’s no standardized consumer tested form that explains key limits of the policy or ways in which the insurance you bought deviates from the standard form. We can’t even get these policies online.  They claim it’s a copyright violation to put the policy online.  (I bet someone could find representation to fight back against this.)

Regulations should require insurers to reveal their policies; should require actual comparisons of prices (the current sites are proprietary and tend to favor one insurer or are lead generators). Regulators’ attempts to control prices create distortions that could be limited by disclosure of what’s actually covered and for how much.  Protects vulnerable groups.

Why have we ignored this fundamental tool of consumer protection in insurance? Path dependency; transparency requires economies of scale that are hard to find on the state level; state regulators don’t want transparency that would show they’re doing a crappy job at command & control; general lack of consumer awareness.

Propose: extend CFPB’s jurisdiction to include insurance.  Would create good regulatory competition: pressure on states to get their acts together.  Threat of federal preemption is the only way to make state agencies act.

Willis: When Nudges Fail: Slippery Defaults (recommended reading!)

Uninsured motorist coverage: NJ default to limited pain and suffering coverage, 80% stayed with the default; PA default to full, 75% stayed with the default: natural experiment.  What about penalty defaults?  UCC: if you don’t specify a quantity in a sales contract, the default is zero: no good faith reasonable quantity gap-filler.  Policy defaults are hot now: 401(k) pension plans as opt-outs are wildly successful in increasing participation rates.  Checking account overdrafts; credit card overlimit fees; etc.  Privacy: web use tracking settings, social network privacy settins.

Insurance has default rules all over the place.  Why are these defaults hot? They promise something for nothing.  Not much political resistance either because everyone can still choose.

Why do defaults stick?  You can opt out of T-Mobile arbitration/class waiver, but it’s in the contract and nobody knows about it: people don’t read contracts.  Invisibility of opt-out options; confusion.  Also loss aversion and the endowment effect.  Procrastination; discounting over time; omission bias (you think you haven’t made a decision yet, though you actually have by not acting) etc.

Preference formation is also important: keeping name on marriage is part of the default, but women opt out in droves—the preferences are clear.  If the signal about what’s expected is very strong, then the default will not stick.

Firm-set defaults work: autorenewal, default privacy settings, autoenrollment.  But rarely noted that the firms also set the altering rules (can make confusing and difficult) and manipulate how it’s framed (if in fine print, you may not know you can opt out). Now though we’re trying to use defaults where firms have interests adverse to the defaults, as opposed to past firm use of defaults.

Looks at overdraft protection: what are firms doing in response to a mandated default that costs them money?  (Overdraft protection is costly; people don’t understand it so competition doesn’t control it; etc.)  Regulators concluded that the best option for most people was no-overdraft for ATM and debit transactions, which should instead be declined.  (Doesn’t apply to checks/automatic payments—people would often rather have the rent check covered than bounce even if they do have to pay a fee.)  Thus that would be the default.  You can opt out though.

Regulation also set altering rules—consumers must take affirmative action to opt out; can’t be hidden in the fine print, and banks can’t bribe you to opt out, providing you the same account terms whether you opt out or not.  And framing rules: must be given information about fees that would be charged; alternatives; must be segregated from other documents.

The situation for defaults seems good: it’s a confusing information environment (banks have weird names for it, like courtesy protection) and there are uncertain preferences—consumers’ preferences are weak if they don’t think they’ll need an overdraft, and conflicted in that they want overdraft protection for emergency situations but not for cups of coffee.  But the default is in fact incredibly slippery compared to 401(k); about 50% of people have opted out.  The people who were overdrafting 10x/year are still doing so because they opted out.

Why?  Banks intentionally undermine each mechanism that would otherwise make a default sticky.  Make opt-out costless and the default costly, for example by sending you gobs of annoying advertising that won’t stop unless you opt-out—calls at dinner; screens before you reach your online banking.  (I have this same issue with PayPal wanting to take money directly from my checking account.)  Neutralize & flip judgment and decision biases, repositioning loss aversion and endowment effect: they say “stay protected” instead of “change.”  Eliminate procrastination with pretend deadlines on opting out.  Hide that it’s a default at all. Chase’s opt-out form makes it seem that there’s no default: it offers you a yes and a no.  Overdraft is effective immediately but no overdraft takes 2 days—this makes no sense because no is the default!  Confusing.  Plus they explicitly advise people to opt out.

Boundary conditions on policy defaults: a sufficiently opposed party with access to the consumer can make the default slippery; the employer is not opposed in the 401(k) situation.  Confusing decision environment; preference uncertainty.  All these can make defaults slippery.

Insurance examples: NY just put in a new default rule: life insurance payment has to be lump sum rather than retained asset account.  Looking at factors, we might think this won’t be sticky: the insurance company is an opposed party with access to the consumer.  The decision environment: may be confusing, may not be.  Beneficiary may focus on the lump sum.  Preference uncertainty may also vary—there’s an endowment effect.  This one might be slippery.

NJ has very high default for personal injury auto insurance.  The point is to allow an apples-to-apples comparison among insurers, but also to create a sticky default.  There’s sort of an opposed party in that the insurer is happy to get the high policy, but also want to be able to quote a lower price for lower coverage, and in fact they are violating the default rule by doing so.  The Ins. Commissioner is trying to come down on this, but might work around it even if complying with the letter of the rule. Decision environment is confusing because people don’t know what their coverage is, and there may be preference uncertainty.

ACPA health exchanges: there is likely going to be an opposed party (the employer might be better off with a crappy policy); we don’t know about access to the consumer (depends on whether you can track the consumer); confusing information enviroment and preference uncertainty are guaranteed.

In questions: it’s incredibly hard to monitor these phone calls and the way they can frame the issue (plus there may be First Amendment constraints on what the regulators can keep them from saying); regulators can only chase behind.  You may need substantive regulation.