Adverse Selection

When one side of a transaction knows more about the sale than the other, the result is adverse selection. The information advantage allows for one participant (Amy) in the trade to benefit from the transaction at the expense of the other party (Betty). The party without the information, Betty, knowing that the transaction likely includes adverse selection, is concerned that the trade may be unfair towards her. Betty may be so worried about that information disadvantage that she will adjust the pricing for that transaction to be more in their favor to her or Betty may simply decide to not do the transaction at all with Amy. If enough transactions for that type of trade are rescinded, then the volume of trades in the market diminishes, which has a greater effect on the market for those transactions in general. In short, distrust leads to inefficient markets.

Insurance Origins

Adverse Selection is a concept that began in the insurance industry. When applying for life insurance, the buyer of the insurance fills out a questionnaire. In that questionnaire, a number of questions are asked that are fraught with misinformative responses (lies). One question would be whether the applicant is a smoker. Another question would be if they are a drinker. Is the applicant a drug-user? How often does the applicant exercise? 

There are limits to the benefit of these questions to the insurance company. First, people have a favorable opinion of themselves. Maybe they normally workout once per week. But there was a week in the last year that they worked out 3 times. They also want to work out three times a week and plan to do so going forward. In their eyes, they are a three-times-per-week exerciser. So they report that they regularly work out three times per week.

Additionally, who answers a question about illegal drug use in the affirmative? Even going into surgery or the emergency room, who would admit that they are a heroin user? No one wants to look bad, even in a confidential survey. 

Lastly and most importantly, someone filling out an insurance form would know very well that if their answers make them seem like an insurance risk, then their premiums would be higher. If you smoke, you pay more. If you drink, you pay more. If you do illegal drugs, you pay much more. If you don’t exercise, you pay more. The only way the insurance company (seller) would know this in an application is by the answers that the insurance buyer provides them. There is no incentive for a buyer to seem riskier. 

Insurance companies know this. While surely insurance companies love their insured, they also distrust them. They don’t distrust them to the point that they simply don’t issue insurance anymore (they would be out of business). Instead, they raise premiums for everyone. What that means is that someone who does not drink, smoke, or take drugs, and additionally works out every day of the week is probably overpaying for their life insurance. On the other hand, alcoholic, chain-smoking drug users who barely leave their couch are likely to pay less than they should for their insurance.

Insurance pricing is more inefficient because of adverse selection.

Market For Lemons

Another commonly cited study regarding adverse selection is The Market for “Lemons” paper by economist George Akerlof in 1970. This study relates to the used-car sales market.

Buyers looking for a car are willing to pay a high price for a good car and a low price for a bad car (lemon). 

Sellers looking to sell their cars are looking for a high price for their good car and would settle for a low price for their bad car. A seller knows whether their car is good or bad. 

If Buyers can determine if a car they are looking at is good or bad, then they will price the car accordingly.  Making that determination for a consumer was nearly impossible in 1970. If buyers thought half the cars that were available to purchase were good and half were bad, then the price they would be willing to pay would be in between those good and bad prices. Assume a good car is worth $15K and a bad car is worth $5K, the buyer would likely pay $10K for any car of that type they are looking at. 

A seller is looking to sell their good car at $15K but they can only find sellers that are willing to pay $10K. The seller of the good car is then not going to do the transaction. Assuming this is the case for all good cars, then the good cars will leave the market of available cars, leaving only a market filled with bad cars, as the study goes. If the used car market is skewed towards bad cars, consumers will be willing to pay less and less and the pricing will continue to deteriorate towards the worth of the bad car. 

The problem, in 1970, was that there was hidden information. No buyers could determine a good car from a bad car by looking at the accident or maintenance history. The used car market was then even more fraught than today with adverse selection.

Obamacare Adverse Selection

Health insurance under the Affordable Care Act (ACA or Obamacare) makes it so that no pre-existing conditions need to be reported and could affect the price of insurance. This provision was intended to make previously very expensive health-care coverage for people with pre-existing conditions to be more affordable. Once insurance is affordable then those people could get health insurance. However, in order for insurance premiums to be truly affordable, then healthy low-risk people needed to pay health-care insurance premiums as well. Under Obamacare, healthcare was mandatory and you would be penalized on your taxes if you did not comply.

Given that there is intentional hidden information, healthcare premiums are more inefficient under Obamacare than previously leading to adverse selection. High-risk people would be underpaying and low-risk people would be overpaying. 

While Obamacare is great for people with pre-existing conditions, as designed to hide incriminating information that would bias premium pricing, it also means that in the end, the pricing for health care insurance assumes that everyone has a pre-existing condition. That resulted in the theory that healthcare premiums would rise across the board. 

In 2017, President Trump pushed through tax legislation that effectively amended Obamacare. That legislation undermined Obamacare saying that healthcare insurance would no longer be mandatory starting in 2019. Those that did not have pre-existing health conditions and were relatively healthy could decide that the premium costs were too high and therefore choose not to be insured and not to be penalized on their taxes. This has occurred over the last few years as 1.2MM people have been added to the uninsured list since it’s low in 2016. 

Number of insured, while lower than in 2010, has been going up since 2016

It should be noted that the people going uninsured are not wealthy health nuts that can self-insure or get really costly gym memberships to help maintain their health. Instead, those uninsured people may generally be healthy but they are mostly low-income people that can’t afford premiums. 

The thought though is that if healthy people leave the pool of insurance payers, then that leaves more unhealthy people, which results in higher costs per insured person which will lead to higher health-care premiums. Adverse selection continually at work…so the theory goes.

Data does not show that there was a jump in the insurance premium cost trend when Obamacare went into effect. Insurance premiums also do not show a substantial jump in the trend in 2019. Still, the growth trend seems to be high as the family coverage cost has gone from $13,770 in 2010 when Obamacare was signed into law to $20,578 in 2019.  Health insurance costs have gone up a lot continuously but blaming Obamacare is more convenient and it is less culpable, so the data shows.

The trend in premium costs has been going up consistently and greatly over time

More Examples

When someone goes to a buffet, the restaurant is betting that the price of the buffet will be fair. However, if one goes to a buffet knowing that they can put down six plates of food, then the cost of the buffet to that person will be a bargain. The individual knows more about their eating habits than the restaurant, who has a flat rate for the buffet. As a result of the restaurant not knowing who is a professional eater, the chances are that the buffet costs more and most people are overpaying compared to what they eat.

A job seeker knows more about their skills and effort than a potential employer. The employer needs to make a decision about A) whether to hire the applicant and B) what to pay them based on very minimal information. It is possible due to this information disadvantage that employers are willing to pay new hires less.

Buying a house has an information imbalance. A seller has been living in the house for many years. A buyer makes a decision to purchase the house based on a limited amount of time seeing the property. That is a big purchase when the buyer is at an informational disadvantage. Chances are that house prices are in fact lower due to this informational imbalance.

When selling an item on craigslist, the seller knows much more about the benefits and likely the faults of the product than the buyer. Items bought on craigslist are cheaper as a result.

On the other hand, some collectors (buyers) may be experts and know more than the ignorant sellers. A collector bought a rare Velvet Underground recording at a sidewalk sale for 75 cents. That record was really worth $25K. The buyer knew more than the seller as can be the case with experts at garage sales and flea markets. At a more formal auction, when the buyers are sure to be collectors, the seller of the item is pressured to raise the reserve price of that item to make up for the informational disadvantage. 

It should be noted that, in all of these cases, when one side of a transaction knows less, then they will adjust the pricing to make it a more favorable risk to enter the trade. A buyer that knows less than the seller will lower the price. A seller that knows less than the buyer does will raise their price. 


In transactions, adverse selection can be a problem for markets as they behave inefficiently. The way to make the adverse selection less impactful is to attempt to remedy the information imbalance. One of those ways is by screening. The product or party is screened which means that it is evaluated through testing or a survey of it to assess suitability for the trade.

In life insurance, oftentimes a nurse needs to do a screening test to take your vitals in order to finalize your application. 

Take Carfax. The buyer of a vehicle now has the ability to see a car’s history and better understand whether it is a good or bad car. Another type of screening in this vehicle transaction would be to take the car in question to a mechanic for a thorough review. 

When purchasing a home, a home inspection, or termite inspection is meant to help the buyer get more information about the property they are about to buy. 

In a job interview, there could be a test or task that the applicant performs to prove they are proficient at the job applied for. 


Another way that the information gap between two parties to a transaction can be narrowed is through signaling. This is when one side volunteers information to help the other side get more comfortable with the transaction. 

When applying for a job, an applicant lists their degree programs and software skills.

In used-car sales, a “certified” pre-owned car has ostensibly been blessed by the auto-dealer as a “good” car. 

In online markets, a frequent seller is ranked by customer satisfaction as a result of their past purchases. A good seller rating makes the buyer comfortable that the seller is legitimate.

A money-back guarantee is a form of signalling whereby the seller indicates that their product/service is so good that if you are not satisfied (find it to be a bad product) they will refund your money. 

Distributors will sell goods to wholesalers. Wholesalers buy a high volume of goods and get a discount for doing so. One has to be classified as a wholesaler in order to get that discount. 

Perhaps the best signaling that I have personally seen is in the watch world (which turns out to be my pandemic obsession though I have never owned a watch in my life). When selling a used watch, if it has “box and papers” then it is considered authentic, eliminating chances that you would be buying a fake. This has a direct impact on pricing. A watch with box and papers sells for a premium of about 20% compared to those without box and papers. That is not to say that you are probably going to purchase a fake if you don’t get box and papers, but that surety has a real value to it. 

Good and Bad Markets

Signaling occurs when:

  • A seller has more information about the product they are selling, believes that they have a good product and if the product is good will raise the offer price of the transaction (as in a used car)
  • A buyer has more information about themselves or the product they are buying, believes that their purchase would benefit the seller, and signals their credentials to lower the price of the offer (as in buying wholesale)

Screening occurs when:

  • A seller has less information about the buyer or the product they are selling, wants to know that the buyer is a good risk, and if the buyer is a good one will lower the price of their offer (as in life insurance).
  • A buyer has less information about the product they are buying, wants to know that the product is good, and ultimately is willing to pay a premium price once it is found to be good (as in a house).

Assuming the product or buyer is good, the seller will signal to raise the offer price of their product and screen to lower the price of their product.

Assuming the product or seller is good, the buyer will screen to raise the price of their product and signal to lower the offer price of their product. 

If the seller, buyer, or product is bad, then it is better from a pricing standpoint to exchange goods on a market that does not have signaling or screening. As those parties cannot pass the test that signaling or screening puts one through in a transaction and would look bad as a result, the product would be more expensive for a buyer on that more efficient market and less lucrative for a seller. By going to a market where there are more unknowns, sellers selling bad products and bad buyers would find pricing more midrange. As an example, if someone wanted to buy or sell a really nice, verified piece of good furniture, they would go to a dealer whereas if they could not prove it was a good piece of furniture, and may even be bad, they will sell it on craigslist.  The dealer for that matter will probably not take the “bad” piece of furniture.

One thought on “Adverse Selection”

Comments are closed.