Will the Number of Uninsured Rise Under ObamaCare?

Published April 26, 2013

It is cruelly ironic, but the massive law that was enacted to solve the problem of the uninsured in America is more likely to worsen it. This would be true even if the program is perfectly implemented and all the provisions come online on time and within budget.

How could this be? It is a multistep process. Stay with me for a second.

The more things change, The more they stay the same.


First, the simplest and most direct form of expanding coverage — Medicaid expansion — is likely to have very little effect. I’m not talking here of the states that refuse to do it after the Supreme Court made it optional, but of the entire program.

Remember that one-third of the uninsured have always been eligible for Medicaid and/or SCHIP coverage but don’t bother to sign up. Actually, it is worse than that. A few years ago, William Sommers wrote in Health Affairs that one-third of all uninsured children had been enrolled in Medicaid or SCHIP within the previous year but their parents found so little of value that they didn’t bother to re-enroll them.

Nothing about ObamaCare’s Medicaid expansion is likely to change this dynamic. Yes, there will be more advertising, and yes a larger number of people will be eligible, but quite of a few of those newly eligible people are already getting coverage on the job, so any expansion of enrollment is likely to be a crowd-out of private insurance. One of ObamaCare’s architects, Jonathan Gruber, has done extensive research on this subject and concluded that as much as 60% of the enrollment in expanded public programs is from people who had been privately insured. No doubt this effect grows bigger the higher up the income scale you go.

By the way, a recent example of this crowd-out phenomenon is revealed in a new study by the Robert Wood Johnson-funded State Health Access Data Assistance Center (SHADAC). Much has been made of the numbers of adult “children” covered under ObamaCare’s mandate allowing people up to age 26 to stay on their parent’s policies. This study shows that the number of such people covered as dependents on employer plans rose from 30.2% of the population group in 2009 to 36.5% in 2011. Sounds like a great success until you realize that the percentage of that age group that had employer coverage in their own names dropped from 21.8% in 2009 to 16.5% in 2011. So, virtually all of the people now covered as dependents were previously covered on their own.

Less studied is the stark reality that many of the people who might be eligible for Medicaid are simply too dysfunctional to enroll. They might be functionally illiterate, drug addicted, mentally ill, outlaws, or in the underground economy and not want to bring attention to themselves. They can’t understand an insurance contract or make and keep appointments for services, but they know where the doctors are 24/7 — the hospital emergency department. When these people have a health problem they don’t need insurance coverage. They need direct care.

Here is where the Supreme Court decision made a very big difference. It said there is nothing illegal about not enrolling in coverage; it simply exposes you to a tax. This removes many of the tools state and local government might have used to compel enrollment. All manner of government services might have been denied to people who do not have proof of insurance — school admission, public housing eligibility, fishing licenses, food stamps, job training, day care — all might have been denied to lawbreakers. But the Supreme Court shut down that possibility. So for low-income people there will be as little compulsion under ObamaCare as there was before and people will continue to behave as they always have.

We can’t calculate what the net effect of all this will be. At best Medicaid expansion will have only modest impact on reducing the numbers of uninsured. But it is equally likely to have no effect at all.

Employment-Based Coverage

Next up is the mind-boggling assumption that employers will continue to provide coverage as they have in the past. No one actually believes this.

In fact, employers have been dropping coverage for at least the past ten years. There is no reason to think this will not continue and may dramatically accelerate under ObamaCare.

The study by the State Health Access Data Assistance Center (SHADAC) cited above finds that the nonelderly population with employer-sponsored coverage decreased from 69.7% of the population in 1999/2000 to 59.2% in 2010/2011. This is because fewer employers offer coverage, and of those that do, fewer employees accept the coverage that is offered. The drop-off is particularly acute for smaller firms with fewer than 50 employees. Only 37.5% of these companies now offer coverage, down from 47.2% ten years earlier.

Once ObamaCare kicks in, many more employers will drop coverage. The only dispute is over how many.

Two years ago the well-respected McKinsey Company conducted a survey of employers and found that 30% said they will “definitely or probably” drop their coverage. The survey was criticized by Obama’s supporters because it wasn’t an economic analysis. Odd, since the same folks seem to live or die according to survey results that are far less rigorous. Avik Roy noted in Forbes that the actual results were even worse than it seemed at first blush. He wrote that the more respondents knew about the law and the more directly involved they were in decision-making, the more likely they were to want to drop their coverage −

…primary decision makers were significantly more likely to drop employee health benefits: 36.5% of primary decision makers said they “definitely or probably” would drop benefits, compared to 22.4% of those who simply had some influence over the decision.

More recently, Douglas Holtz-Eakin, former CBO Director and currently with the American Action Forum, published a study with the economic features the McKinsey critics apparently prefer. He estimated that 35 million American workers will lose their coverage. He assumes most of these will go to the exchanges for subsidized coverage (more on this below), costing the federal treasury an additional $1.4 trillion over ten years.

The Congressional Budget Office is more constrained, but even they have upped their estimate of the number of workers losing coverage from a mere 3 million a few years ago to 7 million just last month.

The CBO number is almost certainly a gross under-estimate. CBO’s ability to predict the future has long been constrained by two things:

  1. It is required to assume that current law will be in effect in the future. So, for example, its budget predictions always assume that the SGR cuts in physician payments will actually occur. But that never happens, so the predictions are never accurate.
  2. It tends to use “static scoring,” which means it assumes that current behavior will be unchanged by new incentives. In this case it issued a 30-page justification for its estimate. As an example, part of that report said −

The fact that many firms currently offer health insurance coverage to their workers despite the high cost of premiums and rapid growth in those premiums for many years shows that many firms continue to find health insurance coverage to be a worthwhile element of their compensation packages. If firms could have attracted employees more cheaply by dropping health benefits and adding wages or other benefits that cost less, then they would have done so.

Good grief! This is about as shallow as you can get. Firms have been offering coverage despite the high cost because there has been no viable alternative, and they feel an obligation to ensure their workers can get coverage. The whole point of ObamaCare is to provide an alternative! Companies will now feel free to drop coverage in the belief that workers will now be able to get good coverage through the exchanges.

On top of actually dropping coverage, no one is estimating the effects of employers who convert full-time workers to part time, reduce the size of the workforce to stay under the mandate, out-source jobs to other companies or even other countries, or enter employee-sharing arrangements with other companies. There is no data for these developments (so they are invisible to policy researchers) but local daily newspapers are awash in stories about companies doing exactly this.

Whether it is McKinsey’s 50 million or so, Holtz-Eakin’s 35 million, or CBO’s 7 million, there is no denying that some large number of workers will no longer have employer-based coverage and will be left to their own devices.

Why should this be the least bit surprising? Kaiser Family Foundation’s 2012 employer benefits survey found that on average employer coverage costs $15,745 per family, of which the employer pays $11,429 (for single coverage it is $5,615, with the employer paying $4,664.) Holtz-Eakin finds that the federal subsidy for a $15,000 plan in the exchange will range from $14,176 for people with incomes of 133% to $2,935 for people with incomes at 400% of poverty ($94,800 for a family of four). For all income groups below 250% of poverty ($59,000), the federal subsidy is far greater than the employer subsidy is.

Employers will be doing their workers a favor if they stop offering coverage, pay the $2,000 fine, and send workers into the exchanges. This is especially true if the company pays out the savings in the form of higher wages. Holtz-Eakin doesn’t even consider the enormous savings for the company if they no longer have to pay the Human Resources cost of finding and negotiating coverage, enrolling workers, explaining the coverage, answering questions, and intervening when there is a problem with a claim. Any CFO worthy of the title would take that trade in a heartbeat.

The Exchanges

So, the Medicaid expansion will make very little difference and some 35 million (perhaps more) people will lose their employment-based coverage. What is left to pick up the pieces? The much-vaunted “health insurance exchanges” (now referred to as “marketplaces” by the federal government).  How will that work out?

Never mind for now the implementation problems (which are massive).  Let’s assume for the moment that they work as planned — they are up and running by October of this year, the hundreds of thousands of newly hired navigators” are competent and well-trained, plenty of insurance companies are participating, and the data-sharing arrangements between employers, state Medicaid programs, and the IRS all work flawlessly. With all of this behind us, what do we have?

Well, first we have the underlying assumption that people really want to have insurance coverage. That is the whole point of this exercise, after all — there are so many uninsured, not because they don’t want it, but because they are deprived of it for one reason or another. One might think somebody would have tested that premise before enacting this boondoggle.

Oops! Three years after enactment, CMS decided to finally ask the question: just who are these poor wretched uninsured people and what are they looking for?

Turns out 92% of them can be divided into three segments:

  1. The biggest cluster (47.8% of all the uninsured) are “healthy and young.” They are not much motivated to enroll and they take their health for granted.
  2. The next largest group (28.9%) are “sick, active and worried.” These tend to be older and are pretty good candidates for coverage.
  3. Finally we have the “passive and unengaged” group (15.3%). I’ve tried to bring some attention to this population (see here). These folks tend to be older and have poor literacy skills.

All of these groups say cost is the main reason they are uninsured, but I expect that is just a throw away excuse. I doubt many of them have the slightest idea what insurance costs. They aren’t interested enough to even look into it.

At best, two-thirds of these people will be hard to reach and even harder to sell (as any insurance agent could have told you years ago.) They are uninsured, not because they are deprived, but because they do not see value in it. The time to do this research would have been before passing the law, not afterwards. That way the law could have been tailored to meet their needs, instead of assuming they will comply with whatever Nancy Pelosi crams down their throats. So out of the 50 million or so currently uninsured we might get 15 million who sign up for coverage.

But what about the newly uninsured, whose employers no longer will offer coverage? Most of these people have been passive recipients of whatever coverage their employers offered. They never had to do anything to secure coverage. We have written about this population before.

For the most part they are very much like their uninsured brethren except they happened to have a job that provided coverage. Once again, one-third may be motivated enough to seek coverage on the exchange, the rest won’t bother, knowing they can always get coverage later on when they need it. Meanwhile, they can save a whole lot of money that would otherwise go the premiums. So, out of 35 million newly uninsured possibly 12 million will get coverage on the exchanges.

But what about the mandate? Won’t that persuade people to get coverage even if they don’t particularly want it? Hardly. The mandate literally has no teeth. The only enforcement mechanism available to the IRS is to confiscate income tax refunds. The vast majority of the uninsured are lower-income (so they pay no federal taxes) and the rest can easily adjust their withholding at the start of the year to avoid sending excess money to the Treasury. No refund = no penalty.

So what are we left with? Medicaid expansion that will enroll few people and most of those will be people who were previously covered (crowd-out). Of the 50 million currently uninsured, possibly 15 million will get new coverage. But these will be offset by the 23 million who lose their employer coverage and don’t bother signing up for exchange coverage. Net result — 8 million more uninsured than before ObamaCare was enacted.

[First published at John Goodman’s Health Policy Blog]