[The] relevance of iPods to the Social Security debate is that this is the rationale for wage indexing rather than price indexing. As general living standards improve, so do one's reasonable expectations of what constitutes a dignified retirement. Today, we have iPods. By the time I retire, there'll be all sorts of new cool shit out there. I shouldn't be stuck, à la price indexing, with 2005 living standards when I'm retired in 2055.
Now, in the column, he makes the additional argument that we should stick with wage-indexing because "that's what the law says" and so we've made an implicit promise to provide those higher payouts. I'm not sure I'd buy that logic in any event, but it takes an extra hit when you recall that wage-indexing was introduced in the late 70s, during an aberrant period when prices were rising faster than wages. In other words, the (myopic) notion was that wage-indexing would save money by resulting in lower benefits. In any event, the fact that the compensation formula changed that recently, combined with the ruling in Flemming v. Nestor, make the notion that there's some kind of implicit guarantee of a specific payout level a hard sell.
That aside, though, there's a problem with Matt's argument that tying benefits to inflation will leave retirees in a kind of time capsule, stuck driving Model-Ts and cranking the Victrola while the rest of us pop music crystals into our flying cars. Now, having an iPod made Rep. John Doolittle (R-Calif.) rethink restrictive copyright laws, so I'm all for cool gadgets for old people. But here's what Matt's missing: As Virginia Postrel has noticed, qualitative improvements aren't typically adjusted for in the Consumer Price Index when measuring inflation. In other words, if a 2005 model car costs more than a 1980 model of the same make, that increase is used to figure inflation. An inflation-based benefit boost, therefore, will already capture the fact that the 2005 model car has all sorts of features—CD-player, power windows, maybe GPS—that wouldn't be standard in 1980. So it's wrong to suggest (as Matt does) that under price-indexing, retirees would be "stuck at the much lower standard of living enjoyed in their youth." Now, since wages do grow faster than prices, a price-indexed Social Security wouldn't raise living standards as much as the current method. But Matt's own arguments about massive technological progress actually undercut the point he's trying to make: Even if they don't enjoy as high a relative standard of living, retirees in 2050 will still almost certainly be considerably better off than their younger-selves today.