Health Insurance by the Numbers — For a Change
Health Insurance by the Numbers — For a Change
The debate about what to do about the Affordable Care Act (ACA) known as ObamaCare is on. This article is about the Individual Health Insurance Market (IM) where people buy insurance who are self-employed, not employed or whose employers don’t offer health insurance as a benefit. I explore how the ACA has affected the IM using publicly available data.
Even though the IM only insures a small fraction of the population, 7%, its economic importance is perhaps more important than its size. The people who buy insurance there are the entrepreneurs starting new businesses. They are people switching jobs to look for more opportunity or a better fit for their talents and lower income folks whose employers do not provide insurance. All are trying to protect themselves against the catastrophic financial loss that a major medical event can cause.
Much of the criticism directed at the ACA involves the IM. Customers are experiencing higher premiums and higher deductibles than they were before. Health insurance companies are losing money in the IM, causing them to leave the market. We hope that the information we provide, which is based on real data and a study of misleading rhetoric, helps citizens, politicians and policy analysts figure out reasonable ways to fix the current problems in the IM.
This article is long but a lot of information is brought together in one place to set the record straight and dispel often perpetrated myths. Given the difficulty of finding data and the inconsistency across data sets, it is easy to get something wrong. If I did, tell me and I will address it.
1. High Cost of New Customers in the IM. The average cost for an IM customer has risen 85% since ACA implementation in 2014 and is now 9% more than the average cost of someone with employer-provided insurance. The average cost of a new customer in the IM after implementation of the ACA is over twice the cost of a person getting insurance from their employer. Unless future IM customers have substantially lower costs, IM average costs will not change much and will keep future premium levels high.
2. Higher Premiums & Insurance Company Losses. Insurance companies in the IM have raised premiums to cover these new expenses in an unsuccessful attempt to remain in the black. By the end of 2015, 2/3 of the Individual State Markets were in the red. However, the combined losses are tiny, on the order of $5 billion. In other words, while the drafters of the ACA surely understood costs would rise, and included subsidies for the Individual Market, they missed providing adequate subsidies — but not by much.
3. Middle Income and Over Customers Most Affected. Since the ACA provided subsidies for low income customers to lower premiums, all other customers experienced the full effect of these high premiums. Subsidies end at $35,000 for single tax filers and $90,000 for joint filers.
4. Premium Disparity with Employer Provided Insurance. There is an enormous disparity in premiums paid by customers in the IM and employees who get health insurance at work. Depending on age and employment status, customers of the IM pay 2 to 8 times what someone with employer-provided insurance pays. Employees benefit from both business and government subsidies while only low income IM customers receive subsidies. The most important factor is the 70% subsidy provided by businesses to their employees. The premium disparity existed before the ACA implementation but is much larger today. Whether intensional or not, the premium disparity is government policy. It is not a market aberration.
5. No Market in Health Insurance. There is effectively no private health insurance market in the US. The private IM represents only 7% of the US health insurance industry while the remaining 94% is a patchwork of federal and state programs including Medicare, Medicaid, CHIP, employer-based health insurance and the Veterans Administration. Given its size, it is not reasonable to think that the IM will be able to move prices into line with employer-based prices — no matter what its structure and terms are. In simple terms, “tails don’t wag dogs.”
Source of Numbers
We will publish a separate piece that explains all the data and estimations. Our sources are the National Association of Insurance Commissioners, The Henry J. Kaiser Family Foundation (KFF), the Centers for Medicare and Medicaid Services, and the United States Census Bureau. We do not say anything that cannot be backed up with this data unless we note an exception. For clarity, we have simplified some aspects of the Insurance Market. Please understand that the numbers are often inconsistent but if I have made any mistakes, I will be happy to fix them.
Problems and Legitimate Solutions
We have a health care cost problem that affects everyone. It’s not just a policy problem or a business or a market problem. It’s also a medical practice problem and a science problem. People are too sick and access so much medical care that the entire society is having trouble bearing the costs. Reasonable people disagree on why this is happening or what to do about it. Depending on your political philosophy and scientific beliefs, you will have different approaches to dealing with this challenge. Nonetheless, everyone can probably agree that this problem will not be solved quickly.
We also have an insurance and healthcare access problem that affects a small part of the population, namely those that use the IM and those that are uninsured. Before the ACA there were high IM premiums and many people who were uninsured and could not access health care, 13% of the population. Three years after the ACA implementation, we have even higher premiums in the IM but less uninsured people. Nonetheless, we still have high premiums and too many uninsured, about 10% of the population.
The high cost of health care is at the root of all insurance and access problems so it is tempting for policy makers to offer to solve the cost problem as a solution to the immediate insurance/access problem. Such approaches cannot succeed because healthcare costs will not fall substantially anytime in the near future. Consequently, these types of solutions are disingenuous at best.
Insurance Sector Size
It is difficult to understand all the IM issues unless you know how big all the pieces of the health insurance industry are. So here is the big picture for 2015. My totals for each sector are for one year. Ten year costs are regularly mentioned in this debate. It is worth considering the motives of people who mention ten year costs only. Sometimes they are reasonable and sometimes their intent is to scare. The numbers will be large no matter what the eventual policies are. Further, It is easy to find sources that differ from these numbers somewhat but no adjustment will change my overall findings and conclusions.
Medicare, a federal government program for seniors, insures 43.3 million people, spends $529 billion per year net of premiums paid by seniors or $1060 per person per month. Many seniors buy additional supplemental insurance privately which we have not included for now in any of our numbers or charts. Medicaid, a state and federal program free for low income people, insures 62 million people, spends $560 billion per year which is $748 dollars per person per month.
Employer-based insurance covers 156 million people, with total spending of about $633 billion amounting to $338 per person per month. About 2/3 of these people work for companies who self-insure and the remaining 1/3 work for companies that purchase insurance. Roughly 20–30% of the premium is paid for by employees and the remaining 70–80% is paid by the employer. In effect, the company portion is actually paid for by its customers — US and even international consumers. There is another critical aspect of the employer-based sector. It depends on a large government subsidy to function. Since a portion of health insurance costs are paid by employers, employees are actually receiving additional income on top of their taxable wages. This additional income, by laws in place since the 1940s, is not taxed. In effect, this tax treatment is a government subsidy.
The IM, a private marketplace now regulated by the ACA, where people choose their own policies, insures 22 million people, at a total cost of $97 billion. Since the ACA, government subsidizes 1/3 the cost of the market or about $32 billion. This subsidy is primarily implemented through ACA Exchanges where people whose income is low enough can purchase products offered in the IM at a discount. The ACA exchanges are part of the Individual Market.
The rest of the people in the country are covered by smaller government programs like the Veterans Administration or they are uninsured. The current estimate of the number of uninsured is about 29 million which is includes low income people of all ages, young people of all incomes, and rich people of all ages who self-insure.
The entire health insurance market is heavily subsidized by either government or business. The subsidies are funded by anyone who pays taxes and consumes — in other words, everybody. The employer-based market tax exemption, depending on average tax rates, is a number that is about $175 billion per year and the business subsidy is more than $450 billion per year. Businesses pass on this $450 billion per year to consumers and it is built into what everyone pays for goods and services. The pre-ACA IM market was not subsidized by government or business and was totally paid for by people in the IM who paid premiums. (The exception were high risk pools, which can be considered to be part of the IM. They were government subsidized in part).
Insured healthcare costs for the under-65 population average about about $5 billion per year for each million people. Using this number, the potential cost for uninsured people for 2015 would be $145 billion per year. Some of these costs were paid, especially by the rich. Others were incurred by the providers of healthcare services, sometimes without payment. If not paid for, providers simply passed on the costs in higher prices to their other customers, again, all of us.
Compared to the entire health insurance industry, the IM is small — some might even say tiny — covering only about 7% of the population. Despite significant regulations imposed by the ACA, it is the only part of the health insurance market where people can choose their own health insurance products. In the employer sector, despite being operated by private companies, employees must pick what their employer offers. Therefore, there is essentially no health insurance market today and absent legislation that abolishes employer-based health insurance, or Medicare or Medicaid, there won’t be a market anytime soon.
The fact that the IM is so small imposes severe economic constraints on the ability of its customers to influence the health care industry as a whole, and, in particular, healthcare prices. Ignoring the scale of the IM has been a favorite rhetorical tactic used by demagogues and political mischief makers from both sides of the political spectrum. Ruining this market will neither deprive most Americans of health insurance nor will increasing its competitiveness allow its limited bargaining power to solve the problem of high healthcare delivery costs now or in the foreseeable future.
Nonetheless, an IM that is disrupted or fails to function, does hurt millions of people — just not enough to swing most elections.
Individual Market and the Uninsured
The combined population in the IM and uninsured people is about 51 million people. Essentially, the ACA tried to address the problem of the uninsured by moving those who could afford to pay something into the IM and those who could not into the Medicaid program that is free to its enrollees. Before the ACA there were 15 million in the IM and 42 million uninsured. Now there are about 22 million in the IM and about 7 million of this population opted for Medicaid. The balance of 29 million remains uninsured, or a little under 9% of the US population of 319 million. The uninsured percent has fallen only by 25% since ACA implementation despite its requirements to obtain insurance.
Today, the IM has about $96 billion per year in claims. Premiums cover about $60 billion of this cost and ACA-based federal government subsidies to the IM provide about $32 billion. Premiums and subsidies fell short of total IM costs of $96 billion by about $5 billion, putting 2/3 of the state-based IM markets in the red. The shortfall is covered by insurance companies’ assets.
At about $5 billion, the total size of the IM’s financial problems are small today assuming the $32 billion in government subsidies to the market remain in some form in the future. If they are removed, the problem will be bigger, at least $37 billion per year. Whatever happens, the problem is still small compared to $1.6 trillion in federal government outlays. Whether government closes this gap in the IM or moves the high cost people to a subsidized high risk market, about the same amount will be spent. If the government only partially subsidizes the cost of these people in the IM, it is effectively asking the people in the IM to shoulder the remaining cost for the national problem of the uninsured. This is the current situation. If government fully pays the cost of the uninsured either in the IM or in a separate high risk entity, then the increased costs would be equitably shared over all taxpayers.
If government does nothing, however, insurance companies must deal with their losses alone. Their options are limited to raising rates or leaving the market, either of which will increase uninsured numbers.
As stated before, the combination of IM premiums and subsidies fell $5 billion short of being in the black, making up about 95% of the total cost of $97 billion. Standard insurance practice for an IM is for premiums to exceed costs by 20% as a cushion for unexpected losses. If the drafters of the ACA had made the reasonable assumption that costs in the post-ACA individual market would approximate the employer market, given the similarity in coverage, their subsidy was way too low. It should have been $19 billion higher to meet industry standards. So, the subsidy should have been 60% higher than the $32 billion.
Whether this was the result of a political bargain or an arithmetic mistake, the result has been to put the IM insurance industry into a quandary. It cannot exclude people or raise prices and deductibles enough to enter the black. Many companies had no choice than to leave the market. That’s the bad news that you read in the media.
The other side of this problem is that the total problem is an extraordinarily small amount of money. Compared to total federal spending of $1.6 trillion, the insurance losses are tiny. To put this in perspective, the gap between insurers making and losing money has the size of a rounding error in the federal budget, about 3/10 of one percent of federal outlays.
Individual Health Insurance Market Costs (Before and After the ACA)
Before the ACA, insurance companies serviced only a fraction of the people in the market today (about 2/3). In order to offer a product at a price insurance companies thought people would find affordable, insurance companies excluded people with pre-existing conditions, discouraged people as they got older from buying insurance by increasing premiums with age, and canceled policies if customers cost too much. Insurance companies could not make money if they let in all the people they excluded because customers would not pay the high premiums. Before the ACA, on a state by state basis, health insurance claims in the IM averaged $198 per person per month using data I analyzed from the National Association of Insurance Commissioners (NAIC).
After two years of ACA enrollment, the IM grew by about 7 million people by the end of 2015. Per capita costs rose to $367 dollars a month, a total increase of 85% since 2012 while the the employer-based sector only went up 12%, roughly 4% per year, about the rate of health care cost inflation.
The drafters of the ACA understood that when the new act did away with pre-existing condition exclusions, costs would likely go up as less healthy people entered the market. They hoped that increases would be partially offset by lower cost uninsured young people entering the market. In order to deal with problem of higher cost customers, Congress enacted multiple subsidies to pay for the increased costs.
Note: Government statistics claim 10–11 million entrants into the IM Exchanges since the ACA was enacted. This is somewhat misleading since only just over half of them were new customers. The others were people already in the IM who qualified for subsidies that they could get in the IM Exchanges.
I did further analysis to determine if people in the IM were spending more on small expenses or whether they were less healthy. For example, people may have been taking care of delayed expenses because they had no fear of cancelation after the ACA. On the other hand, the expenses may have been due to more serious conditions that required hospitalization. This distinction is important as the former is a temporary cost while the latter is a long-term cost that has a good chance of growing. I found that the average number of days spent in the hospital in 2015 increased by 79% from just before ACA implementation. This confirms that the average health of the IM substantially declined. I also estimated that the average cost per month of a new entrant was about $840, more than twice what the average person costs who has employer-provided insurance.
To better understand what was happening, I compared both the costs and the hospitalization in the IM to a large sample of people who received employer-provided health care. Employers do not exclude people with pre-existing conditions nor do they discourage older workers from entering their plans by charging them more. Everybody is charged the same (no reliable statistics available). Given that the ACA made the IM look a lot like the employer market in terms of coverage, unless the people in the post-ACA IM were a whole lot sicker, per-person costs and hospitalization could be expected to be similar in the two sectors.
Before the ACA, the cost per person in the IM was 66% of the cost per person for employer insurance. Days spent in the hospital for the IM were 61% of the hospitalization for those with employer insurance.
As suggested above, the post IM and employer numbers are much closer although the IM numbers are now bigger than the employer numbers. At the end of 2015, the average costs in the employer markets rose to $338 per person per month, putting the IM 9% higher than costs for employers. Hospitalization rose to 21% higher in the individual market. In other words, the average person in the Individual Market after the ACA costs more and was less healthy than a person in the employer market.
On a per-person basis, IM costs and hospitalization are much closer to the employer-provided costs in 2015 than they were in 2012. It is risky to assume that the costs of the remaining uninsured will be similar to the average employer or IM cost today. There is just not enough information to predict their costs yet.
Many have suggested that additional young people in the IM population mix will solve this problem of high costs. These predictions are not likely to come true. While statistics on the age mixtures of the remaining uninsured indicate that young people are somewhat more likely to be uninsured compared to everyone else, the cost of new enrollees are likely to represent the uninsured population as a whole since all ages will enroll, not just young people. So, only if the average cost of the remaining population is far less than the current average will overall averages fall. Further, the experience so far has been that the average cost of new entrants is much higher than the average.
If future policy ends up bringing in more people to the IM that drive average costs higher, there will be additional upward pressure on premiums. It is hard to imagine a scenario where costs will decline. We know what the average cost of employer-provided insurance is for 156 million people or about half the population of the US. If the IM and remaining uninsured look anything like this population, then costs cannot be much different. Given that the remaining population has a very large low-income component and low-income people have higher costs, there is a reasonable chance that costs will continue to go up.
If government subsidies do not keep up with such increases, IM participants will be asked to finance this national problem alone as they are doing now.
Subsidies in the IM
The subsidies in the IM were based mainly on income not age and is the source of much criticism of the ACA. The subsidies stop at incomes of $35,000 for an individual and $97,500 for a family of four no matter what their age is.
Insurance companies are trying to reduce their losses by raising premiums for everyone. The subsidized customers, however, are less affected because they get a break with the subsidies. The unsubsidized members are left with policies priced to handle the losses that are increasingly difficult to afford, if not impossible, the older a customer is.
As we said in the previous paragraphs, the IM was not subsidized enough to pay for the ACA mandates. This is the reason for the sky-rocketing premiums and high deductibles. We should all be asking if the problem of uninsured people should be paid for by members of the Individual Insurance Market alone, especially if they are far less healthy than the population as a whole.
If the costs of these more expensive people were spread over the entire population, the impact on premium prices would be quite small. Insurance is all about sharing of costs among lots of people. If costs are not shared over large enough populations, premiums end up being too high. The existence of uninsured people and people excluded from insurance because of preexisting conditions was and is a national problem. Simple fairness arguments suggest that these costs should be borne by everyone, not a small part of the population.
The ACA provided some subsidies meaning that the general population was contributing some of the cost of the uninsured. However, some is not all and the balance has ended up being financed by by IM members, past and present, not the entire population.
Let’ do the math. The increase in average cost in the IM from implementing the ACA was $167 ($367-$200) per month per person which is $3.7 billion per month. If this were spread over the people in the employer market, this would amount to an increase in their premiums of $24 per person per month. Alternatively, if businesses covered the cost of insuring the higher cost people in the IM, their contribution would go from 70% to 77% of their employees’ cost. This approach amounts to spreading the cost over US consumers for a cost of $11 per month per consumer. The alternative that Congress chose was to have the IM customers bear the burden of admitting higher cost entrants net of the insufficient subsidy, driving up prices and deductibles. Those that did not qualify for subsidies and who were older bore this burden even more.
Whether the mechanism to pay for this problem is taxation or increased prices to consumers resulting from businesses sharing the cost, when everyone shares the cost, it is not a very big problem.
Price Disparity Between IM and Employer-Provided Premiums
Not only are there high premiums in the IM, there is also huge disparity between what people pay for employer-provided health insurance and what people pay in the IM. We have done some estimates to demonstrate IM customers, subsidized or not, pay at least two times what people with employer-provided insurance pay. This is a major factor in the affordability of health insurance in the IM.
There are a lot of variables to consider. Since per capita claim costs are now the same within 10% in both the employer sector and individual markets, premiums would be about the same, absent the various subsidies. However, this is not the case no matter what assumptions are made.
Employers provide a subsidy of about 70% of the cost drastically reducing what people pay for employer-provided insurance. Since the 1940s, employees have paid no taxes on this effective income. Since the ACA, the federal government provides subsidies in the IM market to single customers who make less than $35,000 and families who make less than $90,000. People who make more than these amounts receive no subsidies.
The chart in Figure 5 shows hypothetical ratios of IM premiums to employer-provided premiums for three age categories with and without ACA subsidies. It includes an average age (~40), young age (25) and near Medicare age (62). In all three cases, IM members pay at least about two times what those with employer-provided health insurance pay. When the ACA subsidies are not available, the disparity gets increasingly worse with age.
For the record, there are two variables we have ignored. Self-employed people can take a tax deduction for health insurance and lower their after-tax cost. We have left this out because businesses also get to deduct most of their health insurance costs and I want an apples-to-apples comparison. Even if I include the self-employment tax deduction, the ratios don’t change by that much. I also did not include the difference in average deductible between the IM and the employer sector. The data suggest, but not conclusively, that deductibles are about two times higher in the IM. Including this effect would drive the ratios higher. These variables have similar and canceling effects.
I also explored how the ratios are impacted by the tax free income employees receive from subsidized insurance. Even when I include a tax on that income in the employer sector, the disparity remains significant with the ratio ranging from 1.7–4.0. The reason is that the employer subsidy is the largest factor in determining the disparity.
Large subsidies by businesses provide people who purchase insurance from employers with much lower premiums. Private insurance companies have no way to provide subsidies because they do not have the mechanism that a business has to pass on the cost. While this disparity existed before the ACA, it mostly affected the older than average customers. This problem is now far larger because of the high cost of new customers in the IM.
We are left with only one reasonable conclusion. As long as employer-provided insurance is massively subsidized and the IM has a population with similar costs but with smaller subsidies, premiums will be larger in the IM.
Comments on The Mathematics of Insurance
In order for an insurance company to be viable, it must have enough customers so that its total cost of claims is relatively unaffected by random infrequent high cost events (e.g. long-term hospitalization such as premature birth or major illness). The minimum number of people to achieve this goal can be calculated using fundamental mathematics to be a number in the hundreds or thousands. Most health insurance companies operate with many more customers than this. Even Wyoming, the state with lowest population, does not have this problem, insuring 30,000 people with one insurer.
Once an insurance company has adequate size, its premiums are driven by the average cost of claims per person not the number of people it insures. When cost gets too high, the insurance company has a couple of options. It can deny coverage to high cost people or reduce coverage, options that the ACA outlaws or restricts. Examples include raising deductibles or refusing to cover certain illnesses or procedures.
I include this comment because of various proposals that suggest that increasing the number of people in a group of insured people is a solution to high costs driving high premiums. The mathematics is clear on this point. It is not a solution.
Summary and Conclusions
I have tried to present information that explains what is happening in today’s Individual Health Insurance markets. The numbers paint a pretty clear picture.
Since the implementation of the ACA, a new population of high cost and less healthy people have purchased insurance in the IM. The most likely explanation is that these are dominated by people who had preexisting conditions and who can afford to pay the premiums (subsidized and unsubsidized) offered in the IM. Under the ACA, insurance companies can no longer exclude them. Further, future enrollees from the population of today’s uninsured are likely to cost at least as much people who have employer-provided insurance that comprise about half the US population.
Insurance companies tried to cover the higher costs by raising premiums and deductibles. They were unable to raise them high enough and now 2/3 of the state-based IMs are in the red. Insurers probably did not raise rates higher because they judged that further increases would cause so many policy cancelations that revenues would decline even with higher individual premiums. Still, total losses in the IM are not that big, about 5% of revenues and $5 billion in total. So there is not a large amount of money needed for insurers to get into the black.
These premium increases have now created a large disparity between what people pay for insurance in the IM compared to what employer-provided insurance costs. Depending on age, income and employment status, IM customers pay two to eight times more. This is true for ACA subsidized and unsubsidized customers alike. While this was a problem before the ACA, it is now much worse. While government subsidies have closed the gap some, its contribution is nowhere near the over 70% subsidy provided by employers. Policy makers will have to decide if this disparity is equitable.
So, the small IM is bearing the burden of the nationwide uninsured problem almost by itself. Whether intentional or not, the ACA’s attempt to address the uninsured problem by enabling uninsured people into the IM was not paid for by government. While the ACA attempted to fund some of the additional costs, it did not put enough money into the IM so that it could operate under accepted insurance industry standards where premium revenues are 25% more than claims costs. Insurance companies are simply trying to fix this problem as best as possible by raising rates and deductibles. They have no other source of funding as there is for employer-provided and subsidized insurance.
The IM is too small to absorb a population of higher cost customers without a funding source. Larger insurance companies or groups will not help. As we pointed out in the Mathematics section, the problem of high costs has nothing to do with the IM insurers being too small. The problem is high per-person average costs. Further, these high costs are unlikely to be affected by increased competition in the IM because the IM insures such a small fraction of the total population and the rest of the population is insured by companies not acting in a free market.
The reason why there are so many uninsured people is many fold. People with preexisting conditions often have very high costs compared to what they can afford. High health care costs are beyond the means of low-income and unemployed people. So, if health care cost less, people were less likely to be unemployed and made more money, there would be fewer uninsured. Unfortunately, this will not happen soon and so shorter term solutions must be found.
We are now at a crossroads. Our society needs to decide through its elected representatives whether or not it wishes to deal with the uninsured problem and deal with it fairly. The solution options are more limited than may many might wish. The Employer-provided insurance sector and the Individual Insurance Market are fundamentally different. The employer sector has access to public financing of its subsidies through the sale of its goods and services to the general public. Insurance companies in the IM do not have this option. In the past, insurance companies minimized disparities in pricing between them by insuring only a low cost segment of the non-employer population. Unfortunately, that left out a high cost population with preexisting conditions.
The only way in the short term to minimize the impact of these costs is to spread them over the whole population and that entails government financing. If political philosophy requires that such a solution be rejected, then the decision is to leave a portion of the population without access to health care. These are unfortunately the choices for now.