The a priori use of the consumer confidence argument

I was listening to yesterday's Diane Rehm show and heard the following exchange between a caller named Tina and an economist named Ed Conard.  I've attached Tina's question and Conard's response below so that you can check my characterization of what was said.

Basically, Tina's family is well-to-do but frugal.  Tina asked whether her family's low level of consumer spending is the sort of thing that affects the economy in one way or another.  (Answer:  it slows it down.)

Conard then took the response as anecdotal support for the proposition that the US budget deficit should come down.  He got there by means of what I'll call the US Consumer Confidence argument, which enjoys a good bit of currency among the small-government crowd:

1.  Low US Consumer Confidence is the primary cause of low consumer spending.

2.  High deficits are the primary cause of low consumer confidence.

3.  Therefore, high deficits are the primary cause of low consumer spending.

Perhaps premises 1. and 2. can be proved empirically.  But perhaps they can't, and importantly, the argument is equally logically valid no matter what you assume to be the "primary cause of low consumer confidence" (my language, not his).  For instance, one could substitute "divorce" for "high deficits" and reach the absurd conclusion that divorce is mostly responsible for low consumer spending.  The soundness  of the argument therefore depends on empirical support for proposition 2.

I've generally suspected - perhaps unfairly - that the consumer confidence argument's proponents - let's call them "Republicans" - support premise 2 by consulting their own personal stance on deficit reduction.  That is, "I'm a reasonable person concerned about deficit reduction; therefore other reasonable people must be concerned about deficit reduction, and since most people are reasonable, that means most people must feel a lack of confidence due to the deficit."  Let's refer to this as "projection."

I'm not saying that Republicans are the only ones who project.  After all, I'm not a Republican, and I project, which must mean that every other reasonable person does too.  

The thing is that if you prove the consumer confidence argument by means of projection, then you verge awfully close to assuming your conclusion.  You see that deficits are bad for the economy; therefore everyone sees deficits are bad for the economy; therefore everyone stops spending at the sight of deficits; therefore deficits are bad for the economy.  

In the process, a conclusion about the dangers of deficits is exploded into an apparent causal explanation for how a budget deficit might depress spending.  (Which, of course, is intended as a counterpoint to the conventional macro argument that some or most of the money not spent closing deficits is - you guessed it - spent by consumers.)

Of course, the US Consumer Confidence argument's only tautological if the Republicans who make it are projecting rather than consulting polls and what-not on consumers' feelings.  If the empirical support is there, then, well, perhaps I'm the one who's assuming his conclusion.  

But what's so striking in the exchange below is that you actually see Conard projecting.  Tina says nothing - nothing! - about why her family isn't spending money.  Conrad the budget hawk knows, though.  Tina, after all, is governed by rational expectations.  And "as the deficit grows bigger and bigger, she dials back more and more."