Tim Lee responds to what I've written in the comments of his blog, and here:

But I don't think this is what people are talking about when they say that the mandate is unconstitutional. I think they have something much broader in mind: that Congress shouldn't be using the tax code to force people to do stuff they wouldn't otherwise do and buy products they wouldn't otherwise buy. But if so, then the courts have two options: One is to bite the bullet and invalidate the child tax credit, energy efficiency tax credits, college tuition tax credits, and so forth. Or two, they need a story about why coercing people to buy health insurance is more objectionable than coercing them to have children, pay tuition, take out a mortgage, or install solar panels on their house. Personally, I'd be happy to see the US tax code ruled unconstitutional. But I think it's safe to say that the courts aren't going to do that. And I have trouble imagining a principled argument for invalidating tax incentives to buy health insurance without invalidating a bunch of other tax credits that have long been regarded as constitutionally sound.

I don't think I've claimed that this is unconstitutional--the word has many meanings, but in the end, the only interesting one is "What does Justice Kennedy think?"  I do think it represents a more significant expansion of congressional power than Tim is crediting.

Tim is framing coercion in a strange way to me.  To me a tax subsidy is a subsidy:  you are enticing people to do something by offering them money.  But because this adds the possibility of not getting the money if you don't do whatever they're subsidizing, he's calling this coercion.  This is basically the socialist argument about why the market is coercive, which seems to me to render the word "coercion" useless in discussing liberty; if having more and less attractive options by itself constitutes coercion, then we might as well blow up the universe and start over.

(I think there are distributional justice questions here, but not questions about coercion--except perhaps of the people who are taxed more heavily to make up for any deductions you take).

Let's think about some ways in which we could define this as coercion:

1)  The old socialist way: redefines "coercion" as "living in a universe with scarce goods".  Fun for spicing up dull cocktail parties, but not to me very useful.

2)  "Taxation is coercion":  I see the point but find it ultimately unconvincing.  All taxes have behavioral effects.  Why is the mortgage interest tax deduction more coercive than a cut in marginal rates which may cause people to work more hours?

3)  The government shouldn't try to change your behavior:  Many people might answer the previous question by saying that intent matters:  that the marginal income tax rate change isn't forcing you to work harder, but that this is merely an attractive side effect of cutting marginal rates.  

There's a pretty big problem with this:  things like child tax credits and tuition tax credits are arguably, as I've pointed out before, a recognition of legitimate expenses that should be excluded from taxable income; the list of tax subsidies which are clearly only intended to alter behavior, and do not function to relieve an undue burden on families, is pretty slim.  On the other hand, cuts in marginal tax rates are often explicitly sold as boosting growth by giving people more incentives to work.  So while intent might work as a principle, it does not distinguish the tax credits that Tim lists from cuts in marginal tax rates which most libertarians favor.  Which may be why the Supreme Court seems to have given up on divining intent in most cases.

That is not to dismiss the principle entirely; I think it's right that we tolerate side effects that we would not tolerate as primary effects, precisely because we do not want a government that makes restricting liberties its primary aim.  

I think even the Supreme Court would have no trouble discerning intent here.  But even if it can't, I'd submit the following:  penalties really are different from subsidies.  They are different in their effect on people, different in principle, and different in their effect on governance.  

As I pointed out somewhere, if the government offers to sell you a house for $5,000 which is lovely but has no resale value (it has to be sold back to the government for the price of your capital improvements when you're tired of it), this is economically the same transaction as the government seizing your house by eminent domain, and then offering to sell you a remittance from the seizure for $5,000.  Either way, at time T1, you are poorer by $5,000, have experienced no other change in your net asset position, and have the right to live in a house. 

But these are not the same transaction.  A government that does the former might be wrong, but a government that does the latter is creating political and economic chaos.

(I'm not making such extreme claims about the mandate; it's just a clarifying example.  There are, by the way, programs very similar to the first I described in several cities.  I, for one, am nowhere near as outraged by this as by Kelo; Tim's argument seems to suggest that I should be.)

Here's a way of looking at it that might simplify the debate--or might really confuse it, but here goes.  

Consider "synthetic" financial positions.  These come in all flavors, but here's one of my favorite, because (natch), it involves the government.  The government forces people to save 10% of their income every year for old age.  People respond by lowering other savings, or borrowing money against their forced savings, until their savings rate matches what they were saving before.

In the financial markets, there are all varieties of these, from mimicking stock purchases or sales with collections of options, to more exotic creatures.  During the run-up to the crisis, you had a boom in synthetic instruments, like Synthetic Collateralized Debt Obligations (CDOs), which for example took one side of Credit Default Swaps (CDS) on Mortgage Backed Securities (MBS) and bundled them into new securities--essentially allowing them to create new mortgage securities virtually out of whole cloth.

The first thing I'll observe is that proponents of the argument that there is nothing new here, taxwise, are essentially trying to create a synthetic regulatory penalty, because they're afraid that the court will say they can't just regulate inactivity the way they regulate activity, through the commerce clause.  The taxing powers are arguably much broader--though there are reasons to believe that if the government ever gets as far as needing the tax argument, the mandate is basically done. 

At the same time as they are trying to create this synthetic position (more in their imaginations than in the way that the penalty is actually structured), they are arguing that it isn't any different from things we already do, like tax credits.  This is simply not true, because it's two part, and the tax credit is only one of the parts.  There's also the penalty part. Tim has already conceded that in fact, you probably can't practically replicate the effects of the penalty in the tax code, but he maintains that since we already do the two pieces, he doesn't see why this is in principal an extension of the government's powers.

I would agree that this is not in principal an extension of the government's willingness to nanny us.  But I think that it is an attempt to extend their ability to do so.  And I think we should be very wary of attempts to create synthetic regulations.

In principle, many synthetic CDOs were basically no different from CDOs backed by actual bits of mortgages--either way, you got paid if the mortgages didn't default, and you lost money if they did.  But people overlooked one key way in which CDOs were different--they allowed people to bet money on the mortgage market without reference to the actually underlying supply of houses.  By doing so, they vastly increased the amount of leverage in the system--the percentage of actual assets (houses) underlying bets on housing began to shrink.  The more leverage there is in a system, the more vulnerable it is to collapse.  Arguably, this also blunted signals that investors might have gotten from the market earlier had the CDOs not been available (like, "there aren't really that many houses with mortgages out there").  

In principal, forced savings plus borrowing is the same as no savings at all.  In practice, there can be heavy interest costs, penalties, and even bankruptcies involved.  For example, a lot of people who borrowed money from their 401(k)s lost their jobs when the recession hit, forcing them to either pay off the loan, or pay massive taxes and penalties on their "early cash out"--just when they were least able to afford it.

I think there's also reason to worry about synthetic regulations.  As I've argued at great length, any "synthetic penalty" we could create won't work much like the penalty in the law--it will be more expensive, and less effective.   If it isn't the same in effect, I am unwilling to say, as Tim does, that it is somehow equivalent in principle.  

However, as with synthetic CDOs, the appearance of equivalence might generate a lot of added activity--allowing us to pass laws that we otherwise wouldn't.  Would we have gotten our new health care law with a synthetic penalty that cost more and did less to persuade healthy people to buy insurance?

I'm not arguing that this is the most radical extension of government power since the switch in time, or that we'll all end up living in caves and eating rocks if it goes forward.  I'm just arguing that it's a significant change in both the scope and the means of government power that everyone should worry about unless they genuinely think that there is no reason to care about economic liberty.

We want to hear what you think about this article. Submit a letter to the editor or write to letters@theatlantic.com.