Why the Stimulus Is Not Working

Normally I would twitter this instead of taking the time to blog it but the accompanying graphic is so powerful that it deserves to be blogged. In watching the Sunday talk shows yesterday I noticed that the Administration is once again moving the goal post for how to define success for the $800 billion stimulus package passed in February.

President Obama himself said back then that the measure of success would be jobs “created or saved”, which in itself is a policy game of three card monty because the Federal government does not provide a statistic for “jobs saved”… it’s a meaningless figure if for no other reason than the lack of a historical benchmark to measure relative performance against. Economic statistics are never absolute, the meaning from them comes with comparing to historical performance and in analyzing against related statistics.

On Sunday, however, the Administration and their allies have a new story they want us to chew on, which is that the economy was worse than they realized (more on that in a moment) and that the stimulus worked because it arrested a “free falling economy”. In other words, they saved a cancer patient not because the cancer is in remission but because it hasn’t spread to other parts of the body. On the statistics part of the argument, if the economy was really 2x as bad as everyone thought, what does that say about the Federal governments ability to compile statistical data? Surely in this day and age of real time everything we should expect that the government would be able to compile and distill data and not under or overestimate by 100%.

There is another theory about why the stimulus package isn’t working as promised, Occam’s Razor would suggest that it’s not external factors that are inhibiting positive effects of this package but rather that the package itself isn’t putting taxpayer money to things that will stimulate economic activity. The GAO data below certainly confirms this and it’s even more pronounced when you look at the details for how the money is spent in these programs. Take the billions of dollars given to California for highway and road projects, almost all of that money was passed directly through Caltrans to local governments with no accountability for how the money was to be spent. Caltrans may have as little as $70 million of stimulus funds to throw at projects, not even enough to cover a change order on something like Bay Bridge reconstruction project.

The best explanation I’ve found for why the stimulus didn’t work is this graph from the GAO analysis of the stimulus act. It shows pretty clearly that the 76 percent of stimulus spending through the first four months went to fill in the gaping holes in Medicaid and state budgets. In other words, the stimulus isn’t acting like a pole vault lifting job creation above the baseline. It’s been acting like a crutch to keep state budgets and payrolls from imploding tumbling.

[From Hey David Gregory, This is Why the Stimulus Is Not Working – The Atlantic Business Channel]


More on this topic (What's this?)
Paris march and Islamic extremism: Obama cartoons!
Putin: It Is Time to Play Your Ace in the Hole
Read more on Stimulus Package, Obama's Presidential Policy at Wikinvest

News Analysis, What I Want

200908030828.jpg I don’t want to read all these stories about Eric Schmidt (make that Dr. Eric Schmidt) getting fired from the Apple board of directors, really I don’t. The majority of the stories linked to are linking back to 3-5 source articles and adding a few comments that in retrospect will turn out to not be that unique or insightful.

Here’s what I want: give me a visual map, maybe a hyperbolic tree, of the news topic that segments all of the linked articles into clusters based on sentiment, language quality, uniqueness, and length. While you are at it, throw in a color map that shows how much of the referring article is verbatim text of the source article, which would allow me to skip posts that are essentially copies of the original article, and this would also all me to distinguish which of the source articles is truly a source versus a longer version of the original.