Nate Silver writes about guns. He says:
More elaborate data-mining techniques, such as logistic regression analysis, suggest that gun ownership is a more powerful predictor of whether a voter is Republican than her gender, whether or not she identifies as gay or lesbian, whether she is Hispanic, or whether she lives in the South, along with many other demographic characteristics
Stanley Fish writes about higher education:
The tension between a market model and a Socratic model was nicely captured by two statements Spar made in succession. The first warmed my heart: “We want to teach students things they don’t want to know.” That is, rather than regarding students as consumers (all the rage these days in places like England and Texas), we should regard them as yet-to-be-formed intellects who are often best served by saying no to their desires — as we have traditionally. But then Spar immediately added, “Yet, we can’t be too removed from the marketplace.”
The graph below has been making its rounds (see this post by Tyler Cowen for several links). 
Upon first glance it appears to refute any sort of fiscal approach to restoring GDP. The red (GDP) line is table although the blue (government spending) line decreases. However the key here is noting that this graph actually shows continued growth in GDP once government spending has hit a relatively stable level. It becomes much more clear in the graph below that I put together (note red now represents government spending and blue represent GDP).
Obviously this leaves out many other factors that I would consider relevant to the NGDP recovery (tax stimulus, low interest rates, right-sized asset prices, low steady inflation, etc.), but it at least gives an accurate depiction of the spending-growth pattern that emerged after the recession.
I think he oversimplifies a little when he says:
And the truth is that we know a lot about how to do that — after all, every other advanced country has much lower health costs than we do, and even within the US, the VHA and even Medicaid are much better at controlling costs than Medicare, and even more so relative to private insurance.
The key is having a health insurance system that can say no — no, we won’t pay premium prices for drugs that are little if any better, we won’t pay for medical procedures that yield little or no benefit.
Though the truth is most policy questions have elegantly simple solutions that no one likes.
This is similar to a mistake many commenters make, wanting to distinguish between cash injected into the “real economy” and cash injected into financial markets. Cash doesn’t go into markets at all, it goes into the pockets of people and businesses. There is no meaningful distinction between cash going into the “real economy” and the “nominal economy.” If the Fed buys a bond from a dealer, he’ll quickly deposit the funds in the bank. If the Fed injects cash by paying Federal salaries in cash, the workers will quickly deposit the cash into banks. Over time the demand for cash will rise as NGDP rises. I suppose one could distinguish between cash boosting RGDP and cash boosting NGDP but not boosting RGDP. But then commenters would want to talk about the slope of the SRAS curve, not who gets the money. Or you could talk about cash injections failing to boost NGDP, because the extra money is hoarded. Yes, but once again that depends on factors that have nothing to do with who gets the money, as long as we assume fiscal policy is unaffected.
Jake
Tyler