DISQUS

The only shocking thing about this is that there are actually people who think Terry Gross is a good interviewer. She always asks stupid pointless "gotcha"-type questions that would fit right in on "TMZ". Only instead of these questions going to stupid pointless celebrities, actual thoughtful people with meaningful lives submit to the inanity. The only reason FA has anything of an intellectual patina is that they regularly (not all the time, but much of the time) book interesting and worthwhile guests. Instead of sounding shabby and trite next to her guests, Gross often manages to somehow reflect their intellect without dumbing it down too much. But she never amplifies anything of interest that anyone reasonably familiar with the guest wouldn't already know.

Since she so obviously adores some guests and loathes others (and lets her feelings color every aspect of the interview), my complaints might be written off as those of a crank whose particular idol wasn't sufficiently respected on FA. However, she does an even greater disservice to those she likes. After all, a series of trivial and obnoxious questions at least offer an illusory frisson of the potential for fake conflict, and the way that a guest responds might possibly reflect some of her habits of thought. A series of trivial and fawning questions offer nothing at all. Sometimes you can sense the guest struggling to get across something, anything of interest and significance, only to be repeatedly brought back to earth by some "and how did attending that school at that time affect your impressions of Eastern Europe?" in really-slow-breathy-alto-tones banality.

You don't have to be so abstract: our drug prohibition has created millions of criminals, both nominal (drug possession and distribution) and real (everything else that organized crime entails), both here and abroad. Whenever I hear someone bemoan the sad state of the law that citizens of Mexico enforce in their nation, I feel very uncharitable toward the speaker.

Your critique of Buiter's hypothesized baseline is a little too structuralist for my taste. The noble savages were humans just like us. They had the same intelligence and basic goals, even if they lacked particular customs and institutions. Even more so than for us in the civilized world, any interaction with other parties, including trade, involved a great deal of risk. The tribe down the river might want to sell us some flea-ridden hides, or maybe they just want to see the stuff we have and steal it from us while killing four adults and kidnapping a couple of preteens. Maybe you call that a "transaction cost", but my tribe sees it as more of an existential threat. We'll just shiver through this winter using the hides we have, thanks. Our baseline is to throw rocks at anyone outside the tribe.

Even this sort of prehistoric scenario planning is unnecessary, though. If we really want to do science (that's gonna be a stretch for most economists, but let 'em try), then we're talking about models and hypotheses, not "the truth". It may be easier to work with a baseline of zero than a baseline of epsilon. If the resulting theory allows us to make falsifiable predictions of future observations, then we have something we can use. If not, back to the drawing board. As my own little structuralist interlude above was meant to demonstrate, the less our economic theories depend on fevered aetiological imaginings of the state of nature, the better.

OK, I've seen this jobs-loss graph numerous places now, and I've only just taken the time to read the captions and discover how misleading (or if you prefer, decontextualizing) it is. On average, the total number of jobs goes up over the long term -- just look at the right 2/3 of the graph to see this, or think about population. Therefore, in order to meaningfully compare the deltas of events at different points in time (and different "peak" baselines), we would need to compare their relative magnitudes, not their absolute magnitudes. More conventionally, a graph could just show the total jobs and total losses at all times and allow us to infer what the relative losses were.

As the tiny print in the left margin admits, this is a graph of absolute numbers of jobs lost. That is, the absolute deltas, removed from the context of the peak jobs baselines. That is, we see three slides from zero without being shown how many jobs actually existed at those zero points. That is, this graph doesn't show us anything.

If I just plug some numbers in out of the air, assuming that each successive recession started out with more jobs (which we know to be the case), then the only thing this graph shows is that jobs have been lost more slowly (since the slopes of the deltas are roughly equal for the three successive events, the slopes of the measured amounts must be getting shallower), but for a longer period, in each successive recession. Does this make each successive recession "worse"? Who knows? Not anyone relying on this graph for her information. Since I'm a suspicious person, I'll just go ahead and assume that graphing the relative job losses, or graphing total jobs in a conventional way, would show this recession to be nothing remarkable, which is why those who want us to think the sky is falling created the graph in this misleading way.

All the "division by zero" warnings in your feed are amusing.