Dear Paul,

You might remember me from graduate school at MIT. I would like to ask you a question about what constitutes a reasonable argument.

For example, suppose I were to say, "We should abolish the minimum wage. That would increase employment and enable more people to climb out of poverty."

There are two types of arguments you might make in response. I call these Type C and Type M.

A hypothetical example of a Type C argument would be, "Well, Arnold, studies actually show that the minimum wage does not cost jobs. If you read the work of Krueger and Card, you would see that the minimum wage probably reduces poverty."

A hypothetical example of a Type M argument would be, "People who want to get rid of the minimum wage are just trying to help the corporate plutocrats."

Paul, my question for you is this:

Do you see any differences between those two types of arguments?

I see differences, and to me they are important. Type C arguments are about the consequences of policies. Type M arguments are about the alleged motives of individuals who advocate policies.

In this example, the type C argument says that the consequences of eliminating the minimum wage would not be those that I expect and desire. We can have a constructive discussion of the Type C argument -- I can cite theory and evidence that contradicts Krueger and Card -- and eventually one of us could change his mind, based on the facts.

Type M arguments deny the legitimacy of one's opponents to even state their case. Type M arguments do not give rise to constructive discussion. They are almost impossible to test empirically.

Here are some more examples of issues where liberals could choose to use either type C or type M arguments:

Tax Cuts

Suppose that someone were to say, "The Bush tax cuts will increase long-term growth." You might raise various objections.

One possible type C argument would be that even if the tax cuts increase long-term growth, they will increase inequality. Thus, the consequences are not good. We could have a constructive discussion of that issue, although we may not come to agreement.

Another possible type C argument would be that the tax cuts will reduce national saving, thereby lowering the capital stock, thereby reducing economic growth. They will have the exact opposite of the consequence that is claimed for them. I think that this is an important argument. I have the discomfiting impression that many in the Bush Administration and its supply-side supporters fail to understand this argument.

A type M argument would be, "So what were the Bush tax cuts really about? The best answer seems to be that they were about securing a key part of the Republican base. Wealthy campaign contributors have a lot to gain from lower taxes, and since they aren't very likely to depend on Medicare, Social Security or Medicaid, they won't suffer if the beast gets starved." In fact, this is what you wrote in "The Great Tax Cut Con," which can now be found at The unofficial Krugman archive.

To me, this is not a helpful argument. Imagine that we could somehow prove that the motives of the supply-siders were pure, and that they really did want to improve economic growth. Would that purity of motive outweigh the argument that the higher deficits will actually have the consequence of reducing growth? I would hope not. Conversely, if the motives are wrong but the consequences are good, to me that would argue in favor of the tax cuts, not against them.

School Vouchers

Suppose that I were to say, "I believe that school vouchers would increase the quality of education and reduce the gap between the quality of schools attended by the poor and that of those attended by the rich."

A type C argument would be that there are other values that are more important, so that public education, whatever its flaws, should be maintained as it is. If you took such a position, we could have a constructive discussion, but we might end up having to disagree.

Another type C argument, which you raised in an essay in Mother Jones, would be, "Upper-income families would realize that a reduction in the voucher is to their benefit: They will save more in lowered taxes than they will lose in a decreased education subsidy. So they will press to reduce public spending on education, leading to ever- deteriorating quality for those who cannot afford to spend extra. In the end, the quintessential American tradition of public education for all could collapse." This is an argument about consequences. I believe that it is wrong, because I think that upper-income families would be happy to pay higher taxes to support an education system that works rather than one that fails. But at least we are talking about an empirical question.

A type M argument would be the one you made in the next paragraph of your essay. "The leaders of the radical right want privatization of schools, of public sanitation -- of anything else they can think of -- because they know such privatization undermines what remaining opposition exists to their program." This argument shuts off any constructive debate. It dehumanizes me to the point where I am not even given credit for knowing what my own motives are. Similarly, when I read the comments on Kevin Drum's blog post about vouchers, I see a lot of type M arguments.

The War in Iraq

Suppose that someone says, "The war in Iraq has made us safer from terrorism." You would disagree.

A type C argument would be to suggest that in fact the war in Iraq has made us less safe from terrorism. One view might be that if we had backed down, we would have had broader support in the world and more resources to deal with terrorism elsewhere. It is a difficult question to settle empirically, but we can have a reasonable argument.

A type M argument would be to write, as you did on September 9, that, "It's now clear that the Iraq war was the mother of all bait-and-switch operations. Mr. Bush and his officials portrayed the invasion of Iraq as an urgent response to an imminent threat, and used war fever to win the midterm election." This is not a constructive argument. My opinion is that it requires an implausible degree of complicity among highly dedicated civil servants. Would Colin Powell not have resigned if the purpose of the war were to win an election? Furthermore, I still care far more about consequences than about motives. If the consequences of the Iraq war are that it leads to increased terrorism and conflict, then even the purest motives would not make it a good war, and vice-versa.

The Economic Consequences of Mr. Krugman

Paul, your columns consist primarily of type M arguments. Either you do not see the difference between type C arguments and type M arguments, or you do not care.

I am not going to try to guess your motives for relying on type M arguments. However, I can tell you some of the consequences.

One consequence is to lower the level of political discourse in general. You have a lot of influence with those who sympathize with your views. When they see you adopt type M arguments, they do the same.

Conversely, many of your opponents are stooping to your level. I see type M arguments raised by many of your enemies on the Right. As horse manure draws flies, your columns generate opposition that is vindictive and uninformed.

Another consequence is to lower the prestige and impact of economists. We are trained to make type C arguments. Instead, you are teaching by example that making speculative assessments of one's opponent's motives is more important than thinking through the consequences of policy options. If everyone were to use such speculative assessments as the basis for forming their opinions, then there would be no room for economics in public policy discussions.

You could express your point of view using type C arguments and still take strong stands for what you believe is right. In fact, you might find that doing so would make you more effective. Even if that is not the case, even if there is a sort of media version of Gresham's Law in which specious reasoning drives out careful analysis, then that is a challenge for all of us who are trained as economists. I believe that we have a professional duty to try to be part of the solution, not part of the problem.