The Stimulus Is A Failure

[KrazyHorse]

By the way, it's not simply the absolute scale of fiscal multipliers which is model dependent; even the relative sizes are also not stable across a reasonable spectrum of models.

I have no idea how an analogy to Sharpe ratios works here.

[Vanguard]

I'm having trouble believing that even you are this dense.

When a source presents the Sharpe ratio it says "here is the Sharpe ratio for period X". There is no argument about whether or not what they're presenting is actually the Sharpe ratio. There is a formula for what it is in terms of publically available data. Two different analysts will always report the same Sharpe ratio for the same asset over the same period, as long as they don't make a typo.

On the other hand, a fiscal multiplier is defined in terms of a real versus counterfactual world (comparing overall economic activity to that in a world with a slightly different amount of government spending/taxation). There is no way to directly measure the economic activity of a counterfactual. In order to generate estimators of fiscal multipliers economists plug different policy changes into various macro models that they have. Different macro models generate wildly different estimates of the fiscal multiplier of identical policy shifts. Thus Moody's reporting of a single number to 3 sig figs is either naive or disingenuous.

You epic moron.

This is simply wrong. The figure of $1.73 of economic activity for each $1 of food stamps issued was not generated by imagining how many jobs were saved and calculating backwards.

Kindly try to separate your paranoid fantasies from the topic under discussion.

[KrazyHorse]

How was the 1.73 number generated, praytell?

[Vanguard]

How was the 1.73 number generated, praytell?

I am trying to look up the original, but it may take a while.

[Drake Tungsten]

Good luck getting access to Moody's model.

[KrazyHorse]

Good luck getting access to Moody's model.

I've got it: Moody's found some guy on food stamps and asked him what he would do if he was given an extra dollar. The guy said "buy a 40", so they went to the liquor store and asked the guy at the cash what he would do with an extra dollar. He said "spend 73 cents on a newspaper and flush the other 27 down the toilet".

Economics

[Dinner]

Please tell me you're talking to Lorizael.

[Kuciwalker]

On the other hand, a fiscal multiplier is defined in terms of a real versus counterfactual world (comparing overall economic activity to that in a world with a slightly different amount of government spending/taxation). There is no way to directly measure the economic activity of a counterfactual.

Sliders disagrees with you.

[Vanguard]

Zandi does not explain how the number is calculated, but I still do not see why the fact that his ratios have 3 digits should somehow invalidate them.

The priority of the various stimulus multipliers lies in their ordering, so unless a large number of multipliers should be clustered very close together there is no problem of significance here.

Now, even if somebody were to take the 1.73 multiplier and use it to calculate how many jobs have not been lost in this recession (a very dubious idea, I admit ), you still would not be able to conclude, on the basis of lack of precision, that the "Jobs Saved" number was too high. It might be too low.

In fact, the one good data point we have (The Great Depression) indicates that, if anything, the stimulus has prevented unemployment from going to 30-40%---- not 25-30%.

As for for the whole "counterfactual" argument, I just don't see how it is meaningful. Unemployment would have been some number, stimulus or no stimulus. There is no conditional antecedent present.

[Drake Tungsten]

Sliders disagrees with you.

No. Quinn Mallory and Maximillian Arturo were physicists, not economists, so it's not like the Sliders team was measuring economic activity on the alternate Earths it visited.

In fact, the one good data point we have (The Great Depression) indicates that, if anything, the stimulus has prevented unemployment from going to 30-40%---- not 25-30%.

You are a massive ****.

[KrazyHorse]

Zandi does not explain how the number is calculated, but I still do not see why the fact that his ratios have 3 digits should somehow invalidate them.

Are you honestly this dumb?

[KrazyHorse]

As for for the whole "counterfactual" argument, I just don't see how it is meaningful. Unemployment would have been some number, stimulus or no stimulus. There is no conditional antecedent present.

I have no idea what this is. Some sort of philosophy?

[Vanguard]

Zandi does not explain how the number is calculated, but I still do not see why the fact that his ratios have 3 digits should somehow invalidate them.

Are you honestly this dumb?

Do you ever answer questions, KrazyHorse? Or is your expertise limited to calling the other guy '****'?

[Kuciwalker]

Significant figures by definition represent the amount of confidence in the result. If we know that we can't know a result with more than a certain amount of confidence, additional sig figs are completely meaningless. The fact that you are using 3 sig figs for a number whose measured value varies by up to an order of magnitude depending on the study shows that you're either intellectually dishonest or clueless.

[KrazyHorse]

When I'm dealing with ***** I call them *****.

The 1.73 number is what falls out of some model that this guy has. Fine. If he wants to present his model's estimate to 3 sig figs that's fine too. But:

1) He should have quoted some sort of confidence range, EVEN FOR THE ESTIMATOR FROM A SPECIFIC MODEL. His number is totally meaningless without it.

2) When Moody's as a source claims simply that the fiscal multiplier for food stamps is [blah] and simply writes down the number then it is implicitly accepting the model used by this guy as THE MODEL, PERIOD. As a reporting organization they have the responsibility to assess the state of the art as currently practiced. Instead, they simply took a single number generated by a single model. There is a huge range of plausible fiscal multipliers which come out depending on the model you use. Saying "the fiscal multiplier is [...]" is highly irresponsible, and either naive or disingenuous (as previously stated). Even if they wanted to be lazy they could have said "our chief economist estimated the fiscal multiplier..." and then put in some weasel words about how other economists estimate the fiscal multiplier as a much different number.



****.