Thanks to the recent Supreme Court decision reaffirming the legality of its premium subsidies, the Affordable Care Act keeps on rolling along. Millions have affordable coverage which they highly value, the uninsured rate is the lowest on record, and cost containment appears to be working in ways that are helping both households and our fiscal accounts.

Of course, the assault on the ACA will continue, and one forthcoming line of attack will be on the excise tax on high-cost premiums, aka, the Cadillac tax (even Democratic candidate Hillary Clinton, a solid supporter of the ACA, is apparently having second thoughts about the tax).

Well, I’ve got four reasons why the tax, which kicks in by 2018, should remain in place. First, the Cadillac tax is unlikely to hurt the vast majority of policy holders because most people “don’t drive Caddies,” i.e., most of their premium costs are below the cap. Second, the tax creates an important incentive to hold down coverage costs. Third, the tax should move much of the money employers were spending on health coverage into paychecks. Fourth, Obamacare is clearly reducing the budget deficit relative to earlier projections. But it ain’t free, and if we punt on the almost $90 billion this tax is slated to raise over 10 years, we’ll need to either raise it elsewhere or put it back on the deficit.

According to some numbers I cobbled together from various sources, and making assumptions about the growth of health care costs in coming years, when the tax kicks in, it will hit a grand total of 1 or 2 percent of premium costs. Let me unpack that a bit.

The Cadillac tax applies a 40 percent excise tax to health plans that exceed an annual cost in 2018 of $10,200 for singles and $27,500 for families. While a fair number of policy holders—around 5 to 15 percent, depending on assumptions about cost growth—have plans that cost that much, only a small share of those costs exceed the caps and will be subject to the tax.

No question, that share grows over time. If health costs grow 4 percent faster than the CPI, a fair if not conservative assumption, by 2030, about 15 percent of premium costs would be over the thresholds (which grow at the rate of CPI+1 initially and CPI later) if, and this is a big “if,” employers fail to lower the costs of the plans they offer.

In fact, one point of the tax is to incent employers and providers to find ways to lower premium costs, and this will almost certainly occur, implying my 2030 estimates of who gets dinged are too high. To be sure, some of these costs will surely be passed on to consumers as higher co-pays and deductibles. Others will be absorbed by shaving certain benefits, like narrowing the networks that plan members can access. So no one’s claiming that this will be painless.

On the other hand, economists assume, with some evidence, that when employers pay more for their workers’ health plans, they take it out of the wage part of the compensation package, and vice versa. Thus, affected workers who face some cost shifting are also likely to receive higher pay to offset part of that shift.

So, reviewing what we’ve got so far: the tax initially hits almost none of the premium costs above the threshold. The average family plan today costs around $17,000 which means that in this context, most people are driving Fords, not Caddies. Assuming no reductions in the costs of plans, that share will grow, but a) adjustments will surely occur, and b) while some of those adjustments will shift costs to Caddy-driving workers, they will also shift more wages into their paychecks.

Finally, anyone who wants to kill the tax has some explaining to do. Post-SCOTUS, ideological attacks and votes to repeal the ACA are particularly meaningless, as health reform appears to be here to stay. So what’s your alternative revenue raiser?

And as long as you’re rooting around for a substitute, bear in mind that it should be one, like this excise tax, that incentivizes cost savings such that we can continue the impressive record of the ACA on that front. As my colleague Paul Van de Water pointed out, the CBO now projects that federal health care spending will be almost $700 billion less over the 2011-2020 period than projected in 2010, before Obamacare was in place. I’m not saying the savings incentives in Obamacare currently at work in the U.S. health care system are the only reason health costs have slowed so dramatically, but the weight of the evidence suggests they’re helping.

The smarter move is to leave well enough alone and allow this tax to take effect. After decades of fighting for it, we’re finally moving towards a more rational, efficient health care system. Let’s not screw that up.