Just $105B went to "infrastructure" and was highly politicized to the point projects became difficult to find (for example my governor, one of the least popular in the country, refused [billions of] stimulus money for rail because it was Obama's idea). Republicans were looking at any chance to make the stimulus look like a failure since it wasn't their idea while at the same time looking to take any money they could for their districts. Basically dump on the big obvious projects and then take credit for saving teacher jobs and doing minor infrastructure work all the while not mentioning where the funds came from. It's pathetic.
Even being a tiny chunk, the infrastructure portion has been noticeable where I live. There are two stimulus funded housing developments going up within walking distance, the Federal Court House down the block is getting a much needed retrofit, bridges are being repaired, etc etc.
In actuality, the correct answer is that there are a vast supply of projects going undone for lack of money - crumbling bridges, roads, water pipes, sewer pipes, high-speed rail, just to get started - and that if there had been the political will to spend money on those types of manpower-intensive projects, the economy would be in much better shape today. (You get new bridges, smooth roads, AND a working economy.) Instead a vast amount of money was spent, most of it going to benefit a few thousand of the richest people in the U.S., who have dutifully saved their profits in off-shore tax havens, creating no economic benefit.
When I think of economics, I think of the quote by statistician George Box, "All models are wrong. Some are useful." Economic models are more prone to error than mathematical models in science because they are so complex and can't really account for the human element. And ultimately, their usefulness is subject to their ability to be implemented.
Economics is not closely tied to politics. If you ever want to have a look at what real economics is, please go to http://ssrn.com/ and read some economic articles on subjects that interest you. One thing that you will rapidly notice is that politics has almost nothing to do with formal, academic Economics (and this consitutes the great bulk of the activity that goes on in the field of Economics).
It is worthwhile to distinguish between normative economics and positive economics. Positive economics, which constitutes the majority of the discipline, seeks to describe, explain, and record economic phenomena. Normative economics, which is the louder minority, seeks to prescribe solutions to problems by using economic theory.
The problem with modern Economics is that the loudest voices within the field are listened to because of their political viewpoints and not their economic expertise. Dr. Krugman in particular has the habit of mixing normative and positive economics without providing a careful distinction between the two. This is a huge problem, and it makes the entire field look bad. However, what can economists that are not popular do about this? The mass media has very little interest in impact studies of obscure policy changes done in third world countries, so we just keep churning away in the background.
While I would not say that George Box is wrong outright, I would say that this quote from him is not entirely accurate as it could be when applied to Economic modelling. You see, by definition an Economic model is merely a simplification of a real world phenomena. Of course, all simplifications are "wrong" because they are not intended to be "right" they are merely intended to be approximations.
Economic models are more prone to error than Mathematical models and models in the Natural sciences because they are, in essence, brutally simple models of Human behaviour. Their usefulness is not subject to their ability to be implemented, but is is subject to the context of the problem at hand and what the expectations are for the model.
For example, one thing that I did recently was to analyze some time series data to analyze a market for seasonality. The predictions of my model for the next two quarters were not even close to being exact, but the seasonality estimates were spot on. A model's usefulness must be determined within the context of its overall purpose.
The former are less tied to politics, but much more closely to it than, say, physics.
Actually, they're all of a kind. One of my core problems with Keynesianism is that you must treat "government investment opportunities" like any other resource; you've got low hanging fruit, where you put in a dollar and get $4 in wealth back, you've got a series of barely-above-break-even, then a sea of places you can put a dollar and get back $0.75. You can't just assume that because you've spent a dollar under duress you're getting more than a dollar back in value; economics is built on that but the logic only holds when one is freely choosing to spend the dollar because you expect to get more than a dollar back.
Keynesianism isn't about just spending wildly, it's all about spending where you get over-unity results, and even if I believed in the power of a massive centralized authority to correct determine where those opportunities are (which I don't), there's a huge problem with that, which is that we already spent a large number of those dollars on those over-unity projects. We don't hang around in the boom times and let our roads go to pot on purpose so that when the recession comes we have good investment opportunities for stimulus. It is unlikely that that is even a good idea. So the goverment is trying to stimulate into a market where the low-hanging fruit is mostly gone anyways.
Maybe this works in 1930. Maybe this works in the Third World. In a developed country there's [feasible] no way to PRODUCTIVELY spend three trillion dollars. Now... UNproductive spending we can do, but even under the dictates of Keynesianism that's a bad idea! Sucking a dollar out of the economy (as debt or taxes) to get back $0.75 in value is simply a straight-up net loss.
I find the theory that government spending is producing a lot of net loss of wealth dovetails rather nicely with the real economic results we are experiencing, as opposed to the predictions of those who are implementing this policy.
21st century Keynesianism has some enormous leaps of logic in it that I have not been able to follow wherein it is simply assumed that the government can find things to spend that are stimulative. I'm not even sure where this stuff is supposed to come from, even in theory. You can't even just wave around the word "crumbling infrastructure" because you can't just wave a wand and make everybody a construction worker, so you're bounded there.
Every trick was tried, and that only got us a hundred billion or so, so our politicians just kept spending more anyhow. Guess what? It's disastrous, not stimulative, and as far as I can tell, even Keynes would have said so in advance.