Jim’s Take

Will an HRA Really Benefit My Company?

While many employers are now utilizing HRAs (Health Reimbursement Arrangements), there are still some who are worried that it will leave them vulnerable. There are an even larger number who worry about how to structure the reimbursements.

In this article I will give you the results of 65 “plan years’-worth” of experience with such plans.

Reviewing HRA Concepts and Arrangements

First, let’s review the concept and potential structure opportunities of HRAs. The HRA first became widely discussed about 9 years ago. Many brokers initially pooh-poohed them as too risky, but BBI was an early adapter.

This isn’t necessarily because we’re so prescient (although we are a bit prescient!), but rather, because we had a client that had no option but to adopt such a plan.

An HRA is an “arrangement,” not an “account.” It’s simply an agreement between an employer and the employees that stipulates that the employer will reimburse some of the costs of the deductible on the employee’s behalf.

The idea is that the employer saves some premium on EVERY employee, and only reimburses any significant amount for a fraction of them, the ones that incur the most deductible. [The plans are almost always a reimbursement of the deductible, not the copays or coinsurance, but we do have clients who have done both.]

It’s a risk-sharing concept. It’s similar to self-funding except it doesn’t require a TPA to adjudicate the claims themselves because the fully insured carrier does that. If a claim isn’t covered, nothing will be reimbursed by the HRA. So it’s much simpler to administer than a self-funded plan. It’s also a lot less risk. Typically, the deductible, at least in New England, is $2,000-$3,000 on every single-coverage employee and $4,000-$6,000 on every family-coverage employee.

So the employer’s risk is, for example, the number of singles times the share of the employer’s reimbursement plus the number of families times the share of the employer’s reimbursement. If the employer is reimbursing up to $1,000 for each single and $2,000 for each family and there are 10 of each, then the employer’s risk, i.e. the maximum reimbursement liability, is $1,000 * 10 plus $2,000 * 10, or $30,000.

But Just How Much Risk Is There?

Ah, that’s the question.

And I’ll tell you right now that the answer is, “Not as much as you probably think.”

We have analyzed data on 124 “plan years.” So if ABC company had an HRA with us for three years, that would be 3 “plan years’ of data. I’ll use that data to walk you through how you might design your HRA for your company.

29 of those plan years are for companies with 20+ employees, and 95 plan years are for companies with fewer than20 employees. Clearly, the smaller companies have larger variability of results – the larger the number of employees covered, the more “predictable” and stable the claims experience will be. So look at the companies that most closely match your size of employment.

Moreover, you can choose what share of the deductible you reimburse. Our client fall into three categories:

  • Firms that reimburse the entire deductible. Clearly, the risk is the highest for these companies.
  • Firms that reimburse the first half of the deductible. By deferring the employee’s participation in the deductible, the employer takes more risk on his shoulders.
  • Firms that reimburse the second half of the deductible. This is the most common model. Employees are still exposed to the deductible, so theoretically they’ll be smarter health shoppers, but the employer still minimizes the risk by eating the second half.

Here’s a summary of how they did by employer size:

Company Size Under 20 Employees 20+ Employees
Percent of Deductible Paid by Employer 17.1% 9.4%


Think about that. If you have

  • a group of 20 employees and
  • half of them are single and
  • half are family, and
  • the deductible is $2,000/$4,000, and
  • you pay half,

Then you face a theoretical maximum reimbursement risk of $30,000. That’s $1,000 (half of the single deductible) times 10 singles, or $10,000 PLUS $2,000 (half of the family deductible) times 10 families, or $20,000.

You’re going to pay out, on average, somewhere between $2,820 and $5,130. That will be about 0.9 – 1.7% of premium.

So the question is, “As a percent, how much cheaper is a $2,000/$4,000 deductible plan than a $1,000/$2,000 plan?” Well, it varies by carrier, but the decrement by doubling the deductible is about 14%.

So you save 14% and pay out 1-2%, probably. You’re ahead of the game. By a lot.

Averages Don’t Tell the Whole Story

True. There is some variability there. Here’s how the worst and best cases looked. To avoid weird results, I’ve eliminated employers with fewer than 5 people on the plan. Obviously those tiny plans can easily have a really high year with just a couple of claims.

Company Size Under 20 Employees 20+ Employees
Percent of Deductible Paid by Employer 17.1% 9.4%
Lowest % of Deductible Paid Out 0.0% 0.0%
Highest % of Deductible Paid Out 55.1% 46.6%


Eighty percent of the reimbursement amounts for groups <20 (eliminating those under 5 employees covered) range from 0% - 37.5%. So in the 80% confidence range, the worst case reimbursement of 37.5% means that the employer paid back about 3.5-4.0% of the premium, but they saved 14%, That’s a 3.5:1 ROI. I’ll take that any day.

In groups over 20 employees, 80% of reimbursements ran between 0.0% - 32.8%. Again, something like a 3.5:1 payout.

But It Gets Better

All we’ve compared so far is groups under vs groups over 20 covered employees. Remember, there are three typical ways of reimbursing:

  • Reimburse 100% of the deductible
  • Reimburse the first half of the deductible
  • Reimburse the second half of the deductible.

Reimbursement rates change dramatically when you break it out that way. We have no groups over 20 where the employer reimburses ALL of the deductible, but the difference between first and second half reimbursement is dramatic.

Those reimbursing the first half of the deductible paid an average of 18% of the deductible (vs an overall average of 9.4%), and those reimbursing the second half of the deductible only reimbursed an average of 6.7% of the theoretical maximum reimbursement. The 80%-of-all-results range was 9.4 – 39.5% for first half reimbursers, and 0.0 – 23.6% for second half reimbursers.

For groups under 20, we did have people who reimbursed 100% of the deductible. They averaged a reimbursement of 24.4% of theoretical max with 80% of those amounts coming between 7.4-44.9%. First half reimbursers averaged 19.4% and the 80% range was 0.0 – 42.1%. Second half numbers were 9.8% average with an 80% range of 0.0 – 24.4%

Bottom Line

There are a few conclusions you can draw from this 125 years of experience.

  • 1.An HRA is an almost risk-free program that will generally return you $3.00 for every dollar you risk.
  • 2.It works better if the number of people covered exceeds 20 employees (it’s more predictable, anyhow).
  • 3.It works better if you reimburse the second half of the deductible.

BBI will administer your program at no charge, so there’s no administrative hurdle to jump. All you have to do is ask yourself if you’re willing to take some risk knowing that the average payoff exceeds 3-1, and almost no one ever loses.

We pioneered this tactic more than ten years ago, and it’s been working for some of our clients ever since then. Why not make it work for you. Send me an email – jedholm@bbibenefis.com and tell me you want to evaluate the HRA possibilities.

As Seen In: