The City’s fastest growing expense isn’t in the budget…

The City’s fastest growing budget item doesn’t appear in the General Fund. Council sets the tax ceiling September 8.

There’s a number in Amarillo’s proposed budget that didn’t change this year, and hasn’t changed in two consecutive years.

In a budget where basically everything went up, there it sits, identical to last year, down to the dollar. $23,384,784. That’s what the City contributes to the employee health plan, and it’s exactly what the City contributed the year before.

Normally you’d take that as good news. It isn’t, and the reason it isn’t is the whole point of this post.

Over that same stretch, budgeted medical and dental claims went from $30,758,960 to $34,230,623. Up 11.3%. So claims climbed by three and a half million dollars while the City’s contribution held perfectly still, which raises an obvious question about who paid the difference.

Somebody did. It just wasn’t the General Fund.

Where the money actually went

Amarillo, like most cities its size, is self funded. The City isn’t really buying insurance, it’s paying its own claims, with a stop loss policy behind it for the catastrophic stuff and an administrator handling the paperwork (Aetna). Money flows into an internal health fund from the employer contribution charged out to every department and from what employees pay, and claims flow out the other side.

When claims outrun contributions, the fund covers the gap out of its own balance. That balance is called the “reserve” balance, or just “reserve”.

That’s not a trick, and I want to be clear I’m not accusing anybody of anything except arithmetic. Covering a gap is the entire reason a fund balance exists. A well run health fund carries a reserve, usually around three months of claims, so one ugly year doesn’t turn into a mid year budget amendment and a lot of uncomfortable phone calls. The City’s actual written policy for health plan reserves is: “Three months of claims, or $6M”. Those figures used to be the same. They no longer are, and the City should clarify. Today, “three months of claims” in the medical plan is about $8.5M, not $6M. So the City’s own reserve policy needs to be clarified – is the City going to hold a $6M reserve, or an $8.5M reserve? According to THIS budget, they are aiming at a $6M reserve, which is low. I’ll come back to this point later.

But look at what it does to the document Council votes on.

The General Fund shows the contribution. The contribution didn’t move. So the General Fund shows no health plan increase at all, in a year the plan actually cost $3.5M more. The money was real and it got spent, but it appears nowhere in the fund that drives the tax rate. It’s sitting in an internal service fund summary a couple hundred pages away, in a budget book that weighs four pounds and gets read cover to cover by approximately one person, who is usually the person who wrote it.

Nobody is being deceived here. Council is being shown a completely accurate number that answers a question nobody is asking. Why aren’t they asking? Well, a plausible answer might be that if the budget isn’t calling for “more funding” in the health plan, it is assumed there’s no issue with the health plan at all. In fact, the health plan is fine! Nothing to see here folks….

The cushion, and what happened to it

Here’s the part I find genuinely impressive, and I sincerely mean it.

Two years ago the health fund was carrying about $10.2 million against a $6.0 million reserve target. Call it $4.2 million of reserve surplus, cushion above what policy required. Over the two years since, the City drew $4,444,565 out of it.

Now put those next to each other. Cushion/reserve of $4,204,417. Draw of $4,444,565. The fund ends this year $240,148 under target, which is the difference between the two, almost exactly to the dollar stated in this year’s budget.

That is not a fund quietly bleeding out while nobody notices. That’s a fund walked deliberately down to its reserve floor by people who knew precisely where the floor was. It bought two years of a flat contribution while claims climbed, and it did it without touching the tax rate or anybody’s service level. Honestly, it was the right call.

But it’s also a card you only get to play once.

What happens next year

A cushion is a one time asset. Trend, however, is a permanent condition. It shows up every year, on schedule, like a relative who needs money. “Trend” is how much the health plan’s costs are increasing year over year.

Roll the plan forward one more year at 9% trend, which is my read of the underlying multi year trend and deliberately lower than the 11.3% the City budgeted (this year is actually tracking closer to 12.5, but let’s be generous), and the fund needs about $6.0 million more than it has. Roughly $3.9 million of that lands on the employer contribution, which means the General Fund and the enterprise funds.

There’s no cushion, no more reserve left to soak it up. It has to be funded. And that happens whether or not Council ever touches the tax rate. And it will happen again the next year, and the next year, and so on…..

This is where an invisible line becomes extremely visible, usually at the worst possible moment, because for two years the budget book showed a flat number and everyone reasonably assumed things were handled.

Then it runs into the tax cap

Texas cities can grow maintenance and operations revenue about 3.5% a year before triggering a voter approval election. Status quo health plans in this market are running two to three times that rate (trend). But well-managed plans run under 5% trend.

You don’t need a five year model to see how that ends. You need a napkin.

I built the model anyway. Here’s Amarillo five years out, with the City taking the maximum 3.5% every single year, no exceptions, all five years:

By year five the health plan consumes about 76% of that year’s entire new M&O revenue. Not 76% of the levy. 76% of all the new money the state permits the City to collect, in a year it took every legal dollar available.

Measured against every new General Fund dollar from every source, sales tax and fees and franchise and property tax together, the plan takes 46%, almost half of ALL revenue.

Run the identical five years at 4% trend instead of 9%:

Year five drops to 29% of new M&O, and 19% of all new General Fund revenue. The five year value is roughly $21 million across all City funds, $16.2 million of it General Fund. And by year five the City is spending $5.86 million less every year, permanently. That’s about 30 police officers. Or 2.9 cents on the M&O rate.

For scale, the entire tax relief conversation happening in Amarillo right now, the gap between the proposed rate and the no new revenue rate, is reportedly about $3.4 million. That’s roughly 1.9 cents.

So the health plan trend problem is worth more than the tax debate everybody is currently having. And unlike the tax debate, the health plan issue comes back next year, and the year after that, and the year after that, and the year…..you get the picture.

Why this matters before September 8

Council sets the tax ceiling on the 8th. The rate can come down after that. It cannot go up.

And here’s the mechanic that should bother you. When a Texas city adopts a rate below the maximum allowable, that forgone revenue isn’t actually “gone”. It’s banked for three years under the unused increment, and a future council can go take it back in any of those years without an election.

Which means a rate cut granted against a cost nobody has funded isn’t really a cut. It’s a deferral with a three year fuse. The health plan’s bill comes due well inside that window.

That is not an argument against cutting the rate. I’d cut it. It’s an argument that the cut only survives if the health plan trend comes down with it. Otherwise, we’ve set up some future council to quietly reverse the tax relief and eat the blame for it, which is a hell of a gift to leave somebody.

Four places, and only four

That roughly $6 million has to come from one of four places.

The tax rate. Service levels. What employees pay. Or the cost of the plan itself.

Three of those take something away from somebody. The fourth one doesn’t.

The fourth is also the only large cost the City still meaningfully controls, even if they don’t believe healthcare costs can be better controlled – they can. Debt service is issued. Wages are committed. Sales tax does what sales tax does, and nobody on Council gets a vote on whether people buy trucks this year. The health plan is negotiable every single year, in about a dozen places, most of which never appear on a budget line.

I’ll keep the “how” short, because it’s a different post. On the medical side: direct contracting with local hospitals and physicians instead of taking a discount off a billed charge that was invented for the purpose of being discounted, reference based pricing, steering imaging and infusion and surgery to sites that don’t charge hospital rates for them, centers of excellence, an independent claims audit, a real look at the stop loss placement.

On the pharmacy side, it’s worth saying where the money isn’t. The City’s budget shows $4,941,472 in drug rebate revenue, and by the look of it Amarillo is capturing rebates about as well as a plan this size can. Rebates are the part of a pharmacy contract designed to be looked at. Spread pricing, contract definitions, and the provisions quietly restricting what cost containment a plan is permitted to pursue all move real money and appear on no line item anywhere.

One question

You don’t have to evaluate any of this. I’m a benefits guy with a spreadsheet and an opinion, and you’re entitled to weigh that accordingly.

But somebody should ask one question before the tax rate is set.

What is the projected increase in the City’s health plan contribution next year, and which fund pays for it?

That’s the whole ask. It costs nothing, it takes thirty seconds, and staff either has the number or they don’t. If they have it, Council knows the largest uncommitted pressure on next year’s budget before voting on this year’s rate. If they don’t have it, congratulations, we’ve all learned something considerably more interesting.

The follow up, for a meeting with more room on the agenda: ask staff to bring Council an actual health plan cost containment review, medical and pharmacy both, before the next budget cycle starts. And ask them to put the multi year health fund projection in the budget book somewhere a normal person can find it. The reason this is a surprise at all is that the document shows a flat contribution and never shows the fund running toward its floor. It certainly doesn’t reflect 9% trend.

None of this requires anyone to believe a projection. The trend is already sitting in the claims. The cushion is already spent. The arithmetic runs on its own whether anybody opens the file or not.

The only real variable is whether we find out in September or in March.