President Obama’s signature legislative initiative, the Patient Protection and Affordable Care Act, just may be the worst law passed in America since the Fugitive Slave Act. I say this from personal experience—not with the Slave Act, thank heavens, but with the Affordable Care Act, alas. 

My own family’s health insurance coverage was due to expire on December 31, 2013. We had been paying for it under the provisions of COBRA, the Consolidated Omnibus Budget Reconciliation Act, which allows former employees to choose to continue their employer-provided health insurance, but only for a limited time, usually 18 months. The expiration of our COBRA coverage neatly coincided with the scheduled commencement of Obamacare and the promised availability, on web-based exchanges, of “affordable” and high-quality health insurance policies. But on the basis of my own experience and the reports of many others, the promise remains unkept. The reasons for that failure go well beyond the technical problems that afflicted the national Obamacare website and the state exchanges, as frustrating and amusing as those problems were and are. They are flaws not merely of execution, but of conception, of a law designed to centralize, redistribute, and divide.

If any place should have gotten Obamacare right it should have been my home state of Massachusetts, which pioneered universal coverage under Governor Mitt Romney. Whatever philosophical or ideological problems may have existed with Governor Romney’s idea for the state to force individuals to purchase or enroll in insurance, at least as a technocrat Romney was competent, and by all accounts the Romneycare website, known as the Massachusetts Health Connector, was working fine until fall 2013. Then the administration of Governor Deval Patrick, a Democrat who is such pals with President Obama that he has dinner with the president during Obama’s summer vacations on Martha’s Vineyard, replaced the perfectly functional Romneycare website with a new site, built by CGI Group, the same Montreal-based firm that built the problem-plagued federal Healthcare.gov site. 

CGI had a reported $69 million contract with Massachusetts, but it couldn’t produce a working site. At first I waited, figuring the state website would be fixed on approximately the same timetable announced for the federal site. But problems continued well into December, even as the January 1 deadline approached for enrollment. A notice on the site declared it was closed for “planned maintenance and upgrade activities” from December 11 at 2 p.m. until December 12 at 5 a.m. On December 19, the site was inaccessible again, with a banner announcing “maintenance in progress.” 

When the site did work well enough to at least accept keystrokes, it responded erratically. Once the page I was working on disappeared and was replaced by a notice that said, in part, “???af_dialog.LABEL_OK???” Another time I was told, “Invalid Spouse Name entry.” After hours entering information, it would all be lost: “A connection to the server has failed.”

Don’t just take my word for it: Even the left-leaning Boston Globe admitted in a January 9 news article: “The Connector’s new website has performed so poorly that the state has all but abandoned attempts to fix it.” The paper followed up with a January 15 report: “The deadline for applying is March 24. There is little indication that the website will be working then.”

Eventually I just gave up and got coverage by dealing directly with a health insurance company.

Nor is Massachusetts the only state that experienced such difficulties. Maryland, which, like Massachusetts, has a Democratic governor and state legislature, entered a $107 million contract to build its own exchange. The $175,000-a-year director of the Maryland project resigned in December after taking a weeklong vacation to the Cayman Islands, where she was reportedly unreachable as the Maryland site suffered technical problems.

Why is building a state exchange so difficult? A figure included in the testimony from the secretary of the Maryland Department of Health and Mental Hygiene, apologizing for what he called the “disappointing launch,” provides an idea of the complexity involved. Maryland’s site includes 44 shaded boxes, themselves full of more shaded boxes, themselves full of an alphabet soup of tasks and agencies and technical lingo (“USL/MDM/ETL/IDS”).

These aren’t just websites where anyone can go on and compare prices and features, the way you might go to Amazon.com and shop for flat-screen televisions, or a real estate site and shop for houses, or the Consumer Reports website and compare washing machines or cars. Instead the Obamacare exchange sites want to verify your identity using a Social Security number, address, and birth date. They want to verify your income with the IRS. And they use that information to sort you into three categories—eligible for Medicaid, the government health insurance for the poor and near-poor; eligible for private insurance with premiums subsidized by Obamacare; or eligible to purchase the private insurance at the full, unsubsidized price. For the middle group, the amounts of the subsidies, and thus the net price to the consumer of the subsidized premiums, must then be calculated. 

Under the design of Obamacare, in other words, the centralizing and the class-sorting are indivisible. The program needs to know who you are and how much you earn so it can tell you how much the insurance will cost you. Charging everyone the same price would make for an easier-to-design website—L.L. Bean, for example, doesn’t ask you for your Social Security number or your adjusted gross income before it ships you a pair of duck-hunting boots. The prices are the same for everyone. But the Obamacare sliding scale requires collecting this kind of information.

If a private institution wants to discount its rates along these lines, so be it—plenty of fancy colleges charge high tuitions to wealthy families and offer discounts via financial aid to less-wealthy families, and plenty of private hospitals provide charity care to the poor.

One reason that Obamacare has gotten as far as it has—beyond the interest-group politics of drug companies and insurance companies, which inevitably want more federally subsidized customers—is that many Americans actually agree with the president when he says, in a sense, we are all in this together, and we’re a rich enough country that no one should go without basic health care just because they are poor or lost their job. The tragedy of Obamacare is that a program to carry out the “we’re all in this together” idea is being implemented by sorting Americans carefully into income-rated baskets—poor enough for Medicaid, too rich for any subsidy at all, or just in the sweet spot for subsidized private insurance, with the final category calibrated into large subsidies or smaller ones, with not much accommodation for the possibility that income may vary over time.

Now, one could say that the solution to that problem is a single-payer program or a British-style government national health service. Obama has talked about a gradual transition to such a system, and it may be that in designing the current system with such flaws, he and his allies in Congress were headed in that direction. At some point, even the relatively few left with private insurance may prefer such a system to the alternative of subsidizing everyone else through taxes and ever-soaring premiums.

That, though, has its own risks, not least a decline in the quality and innovation of American health care. 

Whether it wrecks America by dividing us along class lines or it wrecks American health care by turning it into a glorified version of the Post Office, this law is big trouble. The one bright spot is that, just as in the case of the Fugitive Slave Act, a Republican president might eventually be able to repeal it.