Now that the presidential race has started (God help us), candidates have started discussing policies — and because the health care system very obviously needs another round of reform, Medicare-for-all is a hot topic. At a town hall, Kamala Harris was asked whether she would support getting rid of all private insurance companies, and she responded by noting the enormous headache of navigating the private insurance bureaucracy, and concluded: "Let's eliminate all of that. Let's move on."

Chaos ensued. Conservatives attacked, her campaign issued a partial clarification (though she did not disavow her stance), and moderates like David Leonhardt and Jonathan Chait concluded it means Medicare-for-all is doomed. It is a good example of how moderate liberalism can never deliver really quality health-care policy.

Now, it is not strictly true that Medicare-for-all would necessarily mean getting rid of all private health insurance companies. Bernie Sanders' bill, for example, would ban private companies from replicating any coverage provided by Medicare, thus eradicating most of their business model because the new coverage would be so good. However (as is the case in Canada and most similar countries) private companies would probably still provide supplementary coverage for certain things.

The public opinion background here is that while Medicare-for-all polls well on its own, if you tell people it will mean getting rid of private insurance, support plummets from 56-42 in favor to 58-37 against. Thus Leonhardt concludes Democrats should give up on the policy: "I think this particular plan is an unforced error. It comes with huge political vulnerabilities." Chait has a similar recommendation, writing: "If I was advising a Democratic candidate, I would tell them to leave private insurance in place." Instead Leonhardt and Chait both propose a Medicare buy-in as a sort of tactical pre-retreat before the battle has even started.

This is why we can't have nice things.

For one thing, the negative messaging aspect of this poll is not the whole story. The same survey found that if you tell people that Medicare-for-all would "eliminate all health insurance premiums and reduce out-of-pocket health-care costs for most," support increases to 67-30. If you tell them it would "guarantee health insurance as a right for all Americans," it goes up even further, to 71-27.

That demonstrates an elementary political fact about polling: Most people do not have well-formed, ideologically consistent views, and have vague at best notions about policy details. It is thus possible to convince people one way or another through careful choice of wording, context, and framing — that's why "push polling" is a thing.

It follows that a far bigger obstacle than public opinion is the $1 trillion in annual health-care spending that is pure pointless waste — the result of decades of out-of-control cost bloat that is slowly eating the federal budget and destroying the private insurance system to boot. Any health-care reform plan worth the name — whether Medicare-for-all or something else — would radically cut down these costs. Those trillion bucks may be worse than useless in health care terms, but it represents a lot of salaries and a lot of fat profits that can and will be put towards a massive propaganda effort to stop any meaningful reform.

In chemistry, an "activation barrier" refers to the energy needed to start a reaction that works in thermodynamic terms, but needs a kick to get started. (It's why you need a match to start a fire, for example.) Politics often has a similar process at work. To get something good, one usually needs to get over a political hill — that $1 trillion in health-care waste, in this case. From there, one has two choices: Make the policy as good as possible so as to stoke popular enthusiasm to get over the hill or make the policy worse so the hill is smaller. (This is not precisely how the chemistry concept works, but never mind.)

The reason ObamaCare was such a crummy disappointment — with its failure to adequately control cost bloat, coverage of less than half the uninsured population, janky and obnoxious exchanges selling lousy insurance, and regulatory holes — was that it went for the latter option. Instead of confronting the medical-industrial complex, Democrats started negotiations by pre-compromising — lowering their opening bid and buying off almost every interested party.

Chait points to the massive controversy over the ObamaCare implementation period as evidence of the political danger of Medicare-for-all, but this reflects the dangers of timidity as much as it does the dangers of aggressive change. After ObamaCare made it through the legislative process (in which moderates trimmed a bunch more off for no reason) the benefit Democrats had to offer the population through the exchanges was so lame, and the number of people who got it so few in number, that there was not much constituency for the program. (The law's Medicaid extension was a different story.)

Insofar as we care about fixing America's nightmarish hellscape of a health-care system, timidly pre-compromising is not going to achieve much even in political terms. A Medicare buy-in that was any good at all, for instance, would cut deeply into private insurance's market share (because that coverage has gotten so bad and is steadily getting worse), as well as provider cost bloat. Big Medical would likely fight it just as hard as they would Medicare-for-all. And on the other hand, if we start with Medicare-for-all as the opening bid, we might actually get something halfway decent.

Now, I must admit that during the presidency of the guy from Home Alone 2 it isn't always clear which political tactic will work best. But I am quite sure that attacking Medicare-for-all proposals, trumping up political obstacles to its passage, and arguing that Democrats should give up 21 months before the 2020 election is only going to make quality health-care reform of any kind less likely.