Steve Reads

Such is what they wonder at Crooked Timber.

Yesterday I went to a grocery store (don’t worry, this will connect) whose televisions, blaring lights and omnipresent noise gave me a splitting headache that lasted the rest of the day. It’s one of the new super-massive Shaw’s stores, and it competes (as far as I can tell) with only the Stop & Shop across the street.

As it happens, I’ve also been arguing at work with an arch-libertarian and “anarcho-capitalist” who draws his inspiration from Ayn Rand and Murray Rothbard. So I asked myself how my libertarian coworker would defend an incredibly annoying grocery store. To me it’s obvious that there’s nothing to defend: no customer actually asked to have a television assaulting him from directly above the salad bar, and to have another TV assault him in the checkout line. I presume my libertarian coworker would say that if customers really cared, another grocery store would arise nearby that’s not as annoying as Shaw’s, and would steal business from Shaw’s. You could argue that the existence of Whole Foods, with friendlier lighting, pleasant staff and a soothing décor is evidence in favor of this response.

I could retort with a model. I’ve got the outlines of it in my head. Suppose each customer in the area has a utility function that depends on two variables, namely price and “comfort,” where comfort includes the severity of the lighting, the presence and volume of televisions, and so forth. Assume utility is monotone decreasing in price, and monotone increasing in comfort.

As for the stores themselves, assume that they they can offer economies of scale: the larger they are, the lower their price can be.

You can make predictions about which stores will thrive. Of course all will depend on the precise balance of price and comfort in the customers’ utility functions. If price outweighs comfort to a sufficient degree, for a sufficient number of people, the number of potential customers for a new, more-comfortable store will be too small to be financially possible. (The stores must make a certain number of sales just to pay rent.)

This model would give you a plausible story about why there’s a given number of grocery stores in an area. You could falsify that particular story, it seems to me, only if it had any generality: if you then tried to move to another city and apply your price/comfort model to that city, and the model didn’t work, then you’d need a new model. Maybe you’d need to add other variables to price and comfort — variables like per-capita income, accessibility by car and so forth. If your model didn’t work in another city, what you have is an explanation, not a model. It’s not a model unless you can use it to predict something.

Eventually, after enough work on this particular class of model, one hopes that you’d have something that could tell grocery-store owners where they ought to locate. Maybe you’d be able to say that more-comfortable stores would thrive in one place and not in another. I assume there are practical economists working within grocery-store chains who spend all day every day doing just this kind of work.

If you were going to be accurate about it, your model ought to be dynamic: it ought to explain that as more stores in an area turn loud, ugly and assaulting, it becomes harder and harder for people to realize that any choice is available to them. Grocery chains will continue making annoying stores until customers revolt. But customers are growing used to a world where they’re assaulted at every corner by advertising and where shots on television are only a couple seconds long; over time, their tolerance for noise increases and their desire for quiet diminishes. So surely the success of a new, comfortable grocery store will depend on the current stock of grocery stores in the area. It’s not a one-way ratchet — Whole Foods can appear and thrive, after all — but it’s definitely important to factor it into the model.

Making economic models dynamic is the whole point of considering evolutionarily stable strategies, by the way: static equilibria of the sort that you learn in Econ 101 become the endpoints of a process that evolves in time. See also Herbert Simon’s paper “On A Class of Skew Distribution Functions”, which explains the appearance of power laws as the result of a very natural time-evolving process.

All of this involves a good deal more subtlety than the Econ 101-style story that libertarians like to spout. In part, I’m coming to think that this is because they’re confusing the model with the thing being modeled. I heard a libertarian recently say that in a free market, thus and such would happen because there would be infinitely many producers of a commodity product. Which is as ludicrous as saying that such-and-such a thing will happen on the earth because it’s a point mass. “Infinitely many undifferentiated producers of a commodity” is an assumption of a model; it’s not a conclusion. More to the point, it’s an obvious idealization that no sane person would take as a postulate of his worldview.

I do, incidentally, see some value in the political side of anarchism as a first reaction. When you see that the president doesn’t faithfully execute the laws, and the only solution proferred against him is more laws, you wonder if our faith in laws is misplaced. If you actually want to stop dictators, you stop the conditions which make dictators possible. As we all know, a representative democracy like ours is built on a gentleman’s agreement. It’s worth asking whether a gentleman’s agreement is enough to keep dictators away.

So I at least understand the motivation for anarcho-capitalism. When I don’t think that anarcho-capitalists are just mean-spirited and wish to eat the poor, I understand that they actually have the best intentions: they believe that the only way to end the evils of dictatorship is to radically decentralize power. Anarchists only take one step in their analysis, however: they imagine the day after the people have taken back the power that is rightfully theirs, but they don’t imagine the next day when warlords or the Mafia have retaken it. They don’t imagine the feudalism that would (to my mind) inevitably follow. And their critique is wildly ahistorical: capitalism as we know it is only a few hundred years old. To think that the world needs more of this is to offer proof by lack of imagination.

The Olson public-goods book is just  . . .  not very enjoyable. Or at least it didn’t start that way. There’s a particular kind of economic dryness in it, that I can’t exactly pin down with the limited reading I’ve done, but which might be summed up like so: mathematics is used to encapsulate ideas that would be less confusingly presented as prose.

The thing that most people miss about mathematical notation is that it helps clarify things; the best notation actually prevents you from making mistakes. I’m most familiar with probability notation; I’ve clarified a lot of my thinking by properly formalizing what I write. Maybe the same is true with economic notation, and I’ve just not moved far enough along to see that.

So for now I’ve set Olson aside and moved on to an acclaimed biography of John Maynard Keynes. It starts quite promisingly, with a bit of British dryness and mellifluous prose (to combine contrary images for a moment). I’m hopeful.

I watched Josh Beckett absolutely shut down Cleveland’s hitting last night, and watched the Indians’ vaunted ace C.C. Sabathia choke when it mattered. If that was the game that decided whether Sabathia or Beckett would win the Cy Young, it was no contest.

I’m told that the Sox are the odds-on favorite to win the World Series. I don’t know whether that’s true — I know nothing about the NL teams — but if it is, it’s strange: 86 years from the last World Series to the 2004 victory, and then another one just three years later.

P.S. (14 October 2007): Last night’s 13-6, 11th-inning defeat by the Indians makes it clear that this post was somewhat premature.

Over at Crooked Timber, Ingrid Robeyns excitedly describes getting her hands on Amartya Sen’s many-years-out-of-print Collective Choice and Social Welfare. That seemed like a fine thing to read, so I looked in the Minuteman Library Network for it, didn’t find it there, and searched in the confusingly-named Virtual Catalog for it; there it was, and it should be in my hands within a week or so. VirtCat, as far as I can tell, contains MLN; various obscure regulations require that we search within MLN first, and only search VirtCat if the MLN search fails. If the VirtCat search fails, as it sometimes does, we place an interlibrary-loan request with the Cambridge Public Library. I’ve been told that we can email the Ask A Librarian address (reference@cpl.cambridgema.gov) to make ILL requests. I’m not sure how responsive that email address is.

But in any case, I felt like gloating about VirtCat, which can satisfy nearly all the book requests I give it. I need to reiterate that I love living in a town with a thriving library system.

I want to help with their computer problems. The “MLN, then VirtCat, then ILL” request process is somewhat tedious, though I’ve made it less so through the use of Firefox keywords.A while ago I started writing code to combine all these searches in one place, but I’ve not revisited the task in a while. Jessamyn suggested that I get in touch directly with the state library organization; she suggests they’ll be able to get more done than Cambridge itself would.

Finally, I’d like to join the Friends of the Cambridge Public Library, but it’s not clear to me what one gets by joining. I’ve certainly saved more than $500 by using the library, so I’d gladly donate that much. I emailed them to ask whether, for instance, my membership entitles me to help them shape policy; I got no response. Like a lot of Cambridge city government, it wouldn’t surprise me if the library were somewhat behind the times.

It’s funny that a city like Boston, with so many brilliant people and particularly so many engineers, would work as inefficiently as it does. Ah well. Gives me something to focus my energy on. Let’s start with the library.

I’m really impressed what The Onion did with that premise. It could have been another “headline-only” article, but they really made it great. It’s included below.

(more…)

These are my cats, Eve and Lucy (l.-r.). Eve is Lucy’s mom.

That is my first, and I hope only, Friday cat blog.

It’s been some years since I needed to write LaTeX for presentations. (Though I did code up The Babe’s math homework in LaTeX when she was taking a class at MIT last semester; I like to think that she got a couple extra points for that.) Next week I’ll be teaching a little class at work on programming in Python (focused on people who are used to Perl), and I’ll be damned if I’m going to write it in PowerPoint or even in OpenOffice Impress. So I set off to find a good set of templates for LaTeX slides. I seem to recall that back in college, when I last searched for that sort of thing, it wasn’t difficult to do, but it wasn’t trivial either.

Well now it appears to be trivial to do PowerPoint-quality presentations in LaTeX. (See “Blessing, mixed.”) Googling for latex+presentations yields an article on using LaTeX and Prosper. Fumbling around for a couple minutes, I produced my first presentation.

The only gotcha I’d mention up front is that for whatever reason, pdflatex (for generating PDF directly from LaTeX) doesn’t work with this. Instead you have to output DVI first with the latex command, then convert the DVI to PDF with dvipdf.

P.S.: Looking at the source code for the listings syntax highlighter makes my eyes bleed.

 . . . If you so desire. Just click like a good boy/girl if you do.

By contrast to the readiness with which I accept open-source-licensed software, I find that closed-source licenses are an immediate turnoff. I see, for instance, that Adobe AIR is licensed in some way that’s apparently specific to Adobe. Likewise for Adobe Flex.

I think the problem with closed-source licenses is actually very simple and has little to do with ideology; it’s more a matter of transaction costs. When you encounter a non-free license, you as a non-lawyer have to at least give it a quick scan to figure out what you can and can’t do with it. My background assumption as a non-lawyer is that most licenses will be absurd, and won’t let you actually own your software. Maybe it’s a market-for-lemons problem, or maybe that’s just because everything looks like an information-asymmetry problem to me now. The default assumption for someone looking at a new contract is that it will be a lemon, because most licenses are lemons.

By contrast, if you approach a product that says at the top that it’s “BSD-licensed,” or “GPL-licensed,” or whatever, you know what that means. You know, in particular, that you can just start using their code. Actually, to be fair, the GPL presents its own set of problems: if you’re a corporation and you’re looking for software to build on, you may well run away from viral licenses as fast as humanly possible so that the virus doesn’t attach to your code. My layman’s understanding is that this is why a lot of people gravitate toward more-relaxed licenses like the BSD or MIT ones.

There’s no reason why closed-source licenses couldn’t have the same ease of adoption. It’s just a matter of building a specific brand identity around your license. The BSD, MIT, and GPL licenses all have brand recognition; closed-source licenses don’t. If Microsoft, say, published a standard license governing all of its publicly-available libraries, and that license likewise made programmers feel like they could just pick up the code and start writing, they could also build a brand. It hasn’t happened, though.

I’ve started Mancur Olson’s book The Logic of Collective Action: Public Goods and the Theory of Groups. It is most famous, if my outsider’s understanding is correct, for drawing out the free-rider problem: that public goods will be systematically underproduced (or overused) by rational actors, unless there is some means of compelling everyone to participate. Maybe the nearest example is public radio: if we all donate to NPR, we’ll get the programming we love, but any one of us may well think, “Everyone else will donate, so I don’t need to.” Everyone thinks this way, and NPR goes off the air.

This is, specifically, a rational way to think, in the following sense. Suppose I can choose to donate or not to donate. Whether or not I donate, I assume that NPR will still exist (assuming that they’ve not announced their impending bankruptcy on the air). I’ll get the same benefit whether or not I donate. But if I donate, I’ve incurred a cost. Since donating or not donating yields the same benefits, and the former costs more, I’ll do the latter. Everyone thinks the same way, and voilà: no NPR.

(Actually, more specifically you might expect that they’d go through waves of wealth and poverty: when they announce their impending bankruptcy on the air, everyone donates, which makes them flush, which makes them stop announcing the emergency on the air, which makes everyone stop donating. Enough times through this cycle and you’d expect NPR to constantly announce on the air that they’re bankrupt. But then people might realize that the bankruptcy announcements are a lie, so they’d stop donating or NPR would stop pretending that it’s bankrupt. Round and round we go.)

So before I get very far into the book, I’m confronted with the fact that NPR does get lots of listener funding. As a number of behavioral-econ authors that I’ve read (e.g., and most readably, Dick Thaler) have pointed out, people systematically violate the rational-actor assumption in cases like this. Maybe we’re all Kantians, and we all intuitively act as though our actions would become universal law. In any case, I wonder if Olson or others have good responses to this. Clearly there are some legitimate collective-action problems; others that seem like they ought to be, are not.

I’m just at the beginning of Olson’s book, though, and bits of what he’s writing are throwing me off. First he defines a public good straightforwardly:

A common, collective, or public good is here defined as any good such that, if any person X_i in a group X₁,  . . . , X_i,  . . .  X_n consumes it, it cannot feasibly be withheld from others in that group. In other words, those who do not purchase or pay for any of the public good cannot be excluded or kept from sharing in the consumption of the good, as they can where noncollective goods are concerned.

Any number of goods fit into that definition. Clean air may be the most obvious. Or police protection. Olson writes,

A state is first of all an organization that provides public goods for its members, the citizens.

Again, straightforward. And as I assume he’ll point out (or he assumes we’ve already inferred), public goods of the sort that the state provides are often the sort for which rational people would underpay if they weren’t compelled to pay — things like police protection. (“Everyone else will surely pay up . . . ”) But then Olson continues,

[O]ther types of organizations similarly provide collective goods for their members  . . .  There is no suggestion here that states or other organizations provide only public or collective goods. Governments often provide noncollective goods like electric power, for example  . . .  Still, collective goods are the characteristic organizational goods, for ordinary noncollective goods can always be provided by individual action, and only where common purposes or collective goods are concerned is organization or group action ever indispensible.

That’s not clear to me. A corporation is a kind of organization, but it doesn’t exist to provide collective goods. A corporation exists to coordinate the activity of groups in order to provide a good (public or otherwise) that individuals alone couldn’t make. Or maybe it exists because transaction costs would be too high if the goods were produced by individually contracted individuals, à la Coase. But a corporation’s main job, it seems to me, is not to produce public goods, at least as Olson defined them above.

Perhaps he means that even nonpublic goods like automobiles or computers — the classic sort of thing that an industrial organization produces — would be underproduced by individual actors, and would be underproduced for the same reason that NPR would be underfunded: namely, everyone would believe that everyone else would produce a computer, so no one would do it. That can’t be right: you’d contract directly with me to produce an automobile at a specific price, and on we’d go. I would surely produce the car much less efficiently than General Motors does, but it would be produced. It would be produced in lower quantities, surely, but the reasons for the low quantity are not the same as the reasons why NPR would be underfunded.

Can anyone help me figure out what he means here?

Adam Rosi-Kessel, in the comments to my last kvetch about the T, pointed me to two posts on Universal Hub: one about the T’s lack of concern for the safety of its passengers, despite their frequent proclamations that safety is their number one priority, and another about the MBTA piping Adult Contemporary music into its subway stations. This reminds me of two things:

1. The T makes a lot of noise (literally) about customers being their eyes and ears against “the terrorists.” But when it comes to everyday risks like overheated trains or keeping people stuck in the cold, they’re incompetent. Feel free to draw the appropriate general moral about The War On Terror. (Me, I think a War On Car Accidents would be more helpful.)

2. One of the Universal Hub posts linked above mentions annoying buskers. Nearly all buskers are annoying. I’m reminded of Google lawyer and esteemed scholar William Patry’s post, asking how much he’d have to pay buskers to stop playing. I think of this a lot nowadays — e.g., when I go into Shaw’s and see televisions playing above the cash registers and above the salad bar. (Thanks to Adam Rosi-Kessel for pointing me, a few months back, to the Patry post.)

The T may be in something of a death spiral. They raise fares, so fewer people ride, so the T’s income goes down, so service quality goes down, so fewer people ride, and so forth. The spiral is not bottomless; it can be stopped by improving service quality (duh). What’s keeping the T from working properly is not anything technological, and I doubt it’s even economic (in the sense that the funds available to the T from the state government are insufficient to fund a vibrant mass-transit system). At bottom, it’s almost surely political. Maybe the T employees’ union, for instance, has forbidden work rules that would improve service quality. Or maybe it’s even lower-level and more petty: maybe MBTA general manager Dan Grabauskas (who supposedly worked wonders at the RMV) and someone else are feuding. Organizations have been known to grind to a halt for dumber reasons.

I’d love to sit down with Grabauskas and ask him, off the record, why the T works so poorly nowadays. Maybe I’ll email him, naïvely, and ask for a meeting.

I’m sitting in North Station waiting for my 6:45 train to Newburyport, because half an hour is apparently no longer enough time to get from work (an 8-minute walk to Kendall Square) to South Station; hence I missed the bus north from South Station. I’ve missed a few trains/buses by now because of this. The T is decidedly sucktastic recently, as my friend Jason has noted.

I’d like to relay the following conversation with an MBTA employee, but first I’d like you to take a look at the map of Boston’s commuter rails. Note particularly the Newburyport/Rockport line: if you start at North Station and pick the rightmost north-facing branch, that’s the Rockport line (because it ends in Rockport). At Beverly Depot, that line branches off and heads to Newburyport; that’s the Newburyport line. When you get on a train, it goes to one or the other, but not both.

And now to our conversation.

Me: The sign up there says that the 6:45 train goes to ‘Newburyport/Rockport’. Which one does it go to?
MBTA employee: Both.
Me (after a pause): Both?
MBTA employee: Yeah, it goes to both.
Me (after a pause): Thank you.

I cursed, walked 20 feet away, and asked at another window. There the conversation went like this:

Me: Where does the 6:45 train to ‘Newburyport/Rockport’ go?
MBTA employee: Where are you going to?
Me: Newburyport.
MBTA employee: Just stay on the train. If you were going to Rockport, they’d tell you to get off in Salem, wait for the next train to Rockport, and get on it there. Tonight there will be a train where the sign says ‘Rockport/Newburyport’. If you got on that train, and wanted to go to Newburyport, they’d tell you to get off in Beverly and take the train to Newburyport.
Me: Thanks!

This contains several teachable moments about Boston life:

1. The complete incompetence of many (most?) MBTA employees.
2. The helpfulness and friendliness of many MBTA employees.
3. The habit of completely internalizing an arcane set of rules and shrugging at them (if you’ve lived here a few years), or not even noticing them (if you’re a lifer).

That fails to mention the whole reason I didn’t arrive at South Station on time, namely that the T is falling apart. That also is an important part of living here.

In 15 words or less, explain to me why I should vote Democrat in 2008. The normal form of the argument, it seems to me, is this: “Republicans have messed things up; a party with no spine would do better.”

Without cable television, I’ve been pretty disconnected from baseball all year. I could have listened to WEEI, but I didn’t. I’ve been a bad, bad man.

Entering the playoffs, though, I have no choice but to pay attention. I followed the Sox-Angels game the other day via Google text message, then finally got to go to a bar and watch the Yankees pull it out against Cleveland. Finally they went down to defeat last night, so it’ll be Boston-Cleveland in the ALCS.

I’ve no idea who Howard Bryant is, but he has a pretty decent piece on the end of a Yankees era. He says a lot of the Yankee old-timers will be gone or moved out of the regular rotation, and that the team has to start rebuilding from its farm system.

Now it’s Sox-Cleveland on Friday. I can’t wait.

The Ajaxian blog today linked to an interesting-seeming project from Sun, called The Lively Kernel. Like any such product, I do a quick google to see what their license is, and it turns out that it’s GPLed.

I’m noticing this more and more: people are just open-sourcing things by default. Maybe it’s just the circles I travel in, but I don’t think so; the Ajaxian blog doesn’t seem any more free-software-obsessed than anyone else. It’s just that this is how things work nowadays: you write code, and you open-source it. I don’t know why people are doing it; any number of standard reasons, well-rehearsed here, might explain it. All the most dynamic bits of code on the web are open source; recently I’ve run into jQuery (dual-licensed), the Google Web Toolkit (Apache-licensed), the Yahoo! UI Library (BSD-licensed), Google Gears (New BSD-licensed), and and Dojo (Academic Free License, scorned by the Free Software Foundation). That doesn’t mention most of the rest of the Internet’s architecture: Perl (the Artistic License — GNU-approved since it fixed some vagueness), Apache (Apache-licensed), PHP (Zend- or PHP-licensed, which GNU frowns upon), BIND (BSD-licensed), etc., etc.

I wonder if a lot of us had the wrong idea about how the open-source revolution would happen. Maybe it’ll be the case that OpenOffice will mount a successful head-on attack against Microsoft Office, GNOME against Windows, etc. That seems somewhat unlikely. It does seem likely, though, that Apache will embed itself even more as a fundamental part of the web’s architecture, that Firefox will become part of more and more other components, and that in general as more and more moves to the web, more and more will become open source by default. Open-source becomes a revolution as a process, not as a product. Eventually people don’t even think of it even more: they reach for open source as part of doing their jobs. They may not be consciously aware that they’re reaching for something based on the Gecko rendering engine, say, but they are.

It’s a possibility. Doesn’t seem unlikely to me.

My roommates and I are splitting a Netflix subscription, so I’m back on it after a few years away. First up was Dark Days. First bit of advice: Don’t watch it just before bed. A film about those whom John Tierney called “mole people”, living beneath the streets of New York, fighting with the rats for food, is not dream-friendly material.

The film’s subjects are living a pretty comfortable life, though, all told. They’ve somehow managed to get electricity and gas, and few of those whom director Marc Singer interviewed looked emaciated. A number of them, as you might expect, have drug problems; there’s more than one scene of someone down there smoking crack, though Singer doesn’t linger on them.

The most disturbing scenes for me were the more predictable ones that Singer would have to get to eventually: those times when the homeless scavenge filthy, rotten, dripping food out of garbage cans and pronounce it “not bad,” and when one fellow empties the feces bucket. Singer keeps these mercifully brief, but they’re going to make sleep tonight difficult.

Midway through the movie, we learn that Amtrak, under public pressure, is pushing to evict the homeless from its tunnels. Homeless advocates rush in and help get housing for the hundred or so homeless with federal section 8 funding. Singer follows some of his friends to these section 8 homes, and the movie ends there. They seem happy in their new homes, and one guy even lays out plans to furnish his new room with a couple sectional sofas, a twenty-five-inch television, and an entertainment center to replace his bookshelf. It’s not clear where they’re getting the money for this. It’s not clear how those with drug addictions are getting help. A lot of things aren’t clear.

And that’s what bugs me about Dark Days: there’s a lot missing. Where’s their electric power come from? What kind of communities have formed down there? Who decides when the bucket gets emptied? I assume there are drug-addicted scavengers who make life miserable for the rest of them; how do the others fight them off? We see pairs of homeless people becoming friends, and we see the outcome of one fight down there: Dee’s house burns down, and she’s left crying in the corner. Otherwise everyone seems pretty independent. I would have liked a movie that explored their relationships in more depth. If Roger Ebert is right, Singer’s obsession with the “mole people” made him homeless eventually as well; maybe he ran out of money to make the documentary he wanted.

Ebert, by the way, didn’t watch the same movie I did. He gets a number of facts wrong, even in a six-paragraph essay: many of those living underground have dogs, not cats (one guy shows a photo of his kitties, but it’s not clear that they live down there with him; he may have had them aboveground). And Tito doesn’t cook eggplant; he spends a long while talking about how they prepared eggplant in rehab, but he himself never makes it.

I’ve actually noticed that Ebert gets his facts wrong a lot. It doesn’t normally matter, of course, and his reviews are normally qualitatively right. The first time I noticed this was in his review of Brain Candy: it’s “the nipples of mother hope,” not “hype.” And in his recent review of The Lives of Others, he gets it really egregiously wrong: Wiesler is manifestly not the one coldly asking for his coworker’s name after the coworker makes a joke about Erich Honecker; Wiesler’s disgust with that particular scene is crucial to the rest of the movie. Maybe Ebert’s just not taking good enough notes during the films. More likely is that he watches three or four movies a day and unavoidably mixes things up from time to time. I love his reviews enough that I can’t really bring myself to care that much.

Omnivore’s Dilemma can be summarized very quickly: Michael Pollan eats four meals, and tracks down where they all come from. It is a brilliantly simple conceit, and could only be pulled off well by a writer as gregarious, warmhearted, easygoing and scientifically rigorous as Pollan. He wants to know where McDonald’s comes from, so he goes into a cornfield, follows the corn through cows on its way to becoming beef, and visits the “Concentrated Animal Feeding Operations” (CAFOs) in which they’re slaughtered. He interviews corn farmers. He explains the perverse incentives which have motivated corn growers to produce more and more of the stuff, even when it’s not needed. (The government pays them the difference between some set price and the current market price. Hence farmers have an incentive to produce as cheaply as possible.) This is one of the reasons why we as a nation are growing fatter and fatter.

Pollan takes it a step further, though, making something explicit that had never occurred to me: the fact that our country is so nutritionally faddish, leaping from fruit diets to hourly enemas to high-carb diets to high-protein diets, is a sign of something deeply dysfunctional in our relationship to food. (We call the French diet a “miracle,” but Pollan notes that they don’t. They just consider it good eating and a part of the culture.) Pollan never really figures out why we might have this relationship. The lack of a distinctive national cuisine might have something to do with it, he says, but the end effect is clear: we don’t eat well, and nowadays we’re as likely as not to microwave something and eat it in the car. The family meal has been destroyed, and with it the sense of community that food fosters in healthy societies. Pollan’s writing is meticulous and heartfelt, and it made me desperately want to change the way I eat.

After McDonald’s Pollan paints the bright side of the American meal: places like Polyface Farms that are growing more-than-organic food: food that is completely sustainable and delicious. Cows, pigs, and chickens roam widely on a carefully maintained schedule that keeps the grass growing at the optimal rate. The farm produces almost no waste: every last bit of organic matter feeds the next step in the cycle. It’s something of an agrarian utopia  . . .  and it’s probably completely unrealistic for feeding a nation of 300 million people. Indeed, says Pollan, our nation certainly would have capped out at a much smaller population had we not had industrial farming. (It’s a reasonable counterfactual, but it’s debatable.)

After he visits a self-sustaining farm, Pollan tramps off into the wild to hunt and forage for his own food. Also not sustainable at large scale, but that’s not the point: Pollan is trying to reorient us to what meals are about, and how they’re philosophically and ethically larger than just what’s on the plate.

Pollan’s book has made me want to try being a vegetarian again. My girlfriend used to be a vegan, but has turned around 180 degrees and eats a high-protein meat diet. (Atkins vegans are, I imagine, hard to come by.) So the vegetarian thing might have to wait a bit. Being vegetarian isn’t really the sine qua non in Pollan’s book, though; if anything is, it’s short food chains: knowing where your food came from, using food to support your community, and reducing the amount of petroleum necessary to get it to your door. (If peak oil ever comes, bananas may be history.) Joining a CSA is well within my power, and I intend to do so soon.

If I have any gripe about Omnivore’s Dilemma, it’s small: Pollan is a bit too self-satisfied. At one point he eats a meal in the car with wife and child, driving at 65 miles per hour down the highway in California. I don’t actually believe that he wanted to do that. I can hear him saying to himself, “This would make an excellent story for my newspaper article.” Likewise when he’s reading Peter Singer in a steakhouse. If more of the book seemed like Pollan being Pollan, it’d be perfect.

As it is, it is just about perfect. I intend to buy a copy just to have around to shove into people’s hands. It’s a life-changing sort of book.

And so it continues.

This stuff makes me literally sick to my stomach. No wonder most people would choose to tune out. If I had any self-control, I would too.

Thus spake Dana Perino. Her talk has been flagged as a model of obfuscation, and that it was — in her Lowbrow Evil way that she must have learned from Scott McClellan. Ari Fleischer was the last to do Highbrow Evil. But then there’s this:

The President said that a small number of suspected terrorist leaders and operatives captured during the war had been held and questioned outside the United States, in a separate program that was run by the CIA. The President also at the time said that we were not going to — while we had talked about the people that had been held — people, I should say terrorists that were held, they were then transferred to Guantanamo Bay — that we were not going to tell you every time that that happened.

(emphasis mine)

You’ve got to hear the way she delivers that bolded phrase. It’s chilling.

The bureaucrats have their instructions on how to dehumanize others. They’re just carrying out their orders for PR purposes: we’ll now refer to “terrorists” rather than “people,” even if “suspected terrorist” is closer to the truth — and even, for that matter, if we’d refer to any other suspected criminal by his crime (“suspected murderer”) rather than by the way he carried it out (“suspected gunman”).

Welcome to our government, in the early 21st century.

In his Nobel lecture, George Akerlof notes that workers resist nominal wage cuts, so wage changes are much more biased in the positive direction than strict rationality would imply: a neoclassically rational corporation would cut wages more often than it does.

But I wonder: do corporations actually cut their average wages through attrition? Higher-wage employees retire, and they’re replaced by lower-paid ones. Any individual employee’s wages aren’t affected, but wages for a given task decline. I’d like to read the studies Akerlof cites to see if they’ve taken this into account.

(Insert obvious caveat for employees working under union contract.)