The ATM Myth

Brad DeLong cited this passage from Ron Suskind’s latest book on Monday:

Both [Christina Romer and Larry Summers], in fact, were concerned by something the president had said in a morning briefing: that he thought that high unemployment was due to productivity gains in the economy.

The same meme spread across the economics spectrum: Scott Sumner was horrified. Mike Konczal’s reaction (on Twitter) was restrained (“This is…depressing”) by comparison.

In the context of Suskind’s book, we might just assume that Obama was, as usual, being gulled by his handler Rahm and his Svengali, Timmeh Geithner. However, via Karl Smith, we can set to rest any doubt that Barack Obama is just being misled. Matt Yglesias catches a lazy piece of  “thinking” from the President—who has had Austan Goolsbee, Larry Summers, Peter Orszag, Alan Kreuger, Jason Furman, Jason Bernstein, and several others (even dissing Christina Romer, as the Now-Sainted-by-the-Press-for-his-Fairness-to-Women Obama explicitly did) to correct him:

In a June interview with Fox News, President Obama appeared to argue that the country is suffering from high unemployment because productivity enhancing technologies such at ATMs have reduced the need for work.  It wasn’t clear to me at the time if the president really meant that or if it was just a bad moment in an interview,  . . .

Team Obama has, I think, landed on a more sophisticated version of this theory, and that explains some of the reason why Romer & Summers aren’t in the administration anymore and haven’t been replaced by like-minded people. [link in original]

Now, Barack Obama “might not be particularly well-informed about economics” (Sumner), but I never would have thought he was that obtuse.

Let me talk for a moment about our shared experience. Obama was a year behind me in college. As a transfer student, he didn’t start with on-campus housing; he got a room in a flat in the mid-100s.  So he walked to school—past several banks, one of which was undoubtedly more than happy to open an account for a never-attended-public-school Ivy Leaguer who, even then, knew how to manage up. (Heck, they opened one for me, and I fail miserably at most of that description.)

And the one thing you got—whether it was Chase (back before Manny Hanny acquired it) or Citibank (before it became The Big C)—as a student, no minimum balance required, was an ATM card, usable (in the case of Chase) virtually any time in one of the three (3) enclosed ATMs.  They dispensed $5 and $10 bills.

(The “enclosed” is essential, not just for nighttime safety, but also for days of rain and snowses. Years later, in MBA school, we attended the presentation of a guest speaker, a prominent Georgian who had founded an “Internet bank.” He freely stated that he didn’t understand why some companies enclosed their ATMs; I put the now-bankrupt company on my “short” list immediately.)

That was thirty (30) years ago. Putting this as politely as possible, that’s a heckuva long time for “structural change” in employment to take effect.  To put it in context, I know hot type setters who have been out of that business for less time.

It would take someone who was completely unaware of the world around him to point to the ATM as a ca.-2007 cause of structural unemployment in the United States.

Now, you may, correctly, note that there are Many More ATMs in 2011 than there were when Barack H. Obama moved to Morningside Heights in 1981.  And I will certainly agree with you: after all, if you spend 25 years privileging capital investment over labor—not to mention thirty attacking labor—you should expect capital growth.

But that doesn’t mean that capital growth is welfare-enhancing, or even necessarily has a positive ROI.

At Bear Stearns’s main building, 383 Madison, there were two Chase ATMs on the second floor. Those who know NYC will note that there are not one but two Chase branches across the street, but traders and senior executives require convenience, not crossing a NYC street (which they only did to cash/deposit their bonus checks) or, especially, standing in a queue behind people whose time is much less valuable.

(Tone notwithstanding, the last part of that is rather serious: if being away means you or your firm might take a loss in the mid-five or low-six figure range, your work time is more valuable than that of a Claire’s Stores saleswoman. Much better when the worst-case scenario is to be stuck behind one or two of your fellow traders, who wouldn’t waste their time not knowing exactly what they intend to do before getting to the machine.)

And, since the traders and executives weren’t willing to be personally ripped off, the ATMs were no-fee, even for non-Chase cards.

And here is where an alert economist will say cui bono? The answer is: the firm paid $1.2MM ($1,200,000; $600,000 per machine) for those two machines to be there, be maintained, and be fee-free.

So when you point to those ATMs in the grocery store, I point to the 99 cent ($0.99) fee for a $40 withdrawal and say glibly, “Yes, usury is alive and well; you don’t even need to be in a counting house.” 

How many jobs does that ATM destroy?

And—again, if you’re able to think about it, the way some economists (mostly health economists and econometricians, it seems) are able to do—you look at the proliferation of ATMs and realize that each of them needed to be stocked with bills (now $20s and $50s), have their network connections maintained, be repaired, have (a portion of) a staff member available to deal with customer complaints and issues, and be installed based on an initial capital outlay/agreement that covers all of those costs.

If we treat the Bear Stearns arrangement—and Bear and Chase were tight even before the latter “paid the two dollars”—as a benchmark, a semi-full service (as with the ATMs in most grocery stores and Duane Reades, they didn’t take deposits) ATM costs a lot more to run than a bank teller does. As a ballpark, it may be an order of magnitude more, and that’s rounding the ATM costs down (ca. $500K) and the teller expenses up ($50K).

All of the above is even ignoring The Baumol Problem.  You can provide most of your services through an ATM, or a group of ATMs, if it/they work(s) perfectly, but you cannot eliminate the “teller” role completely. You can redefine it: a bonded employee to restock the bills being dispensed is needed, so you can take a teller away from a “window” for an hour a day: seven hours as a teller, one in service of the ATM. And you can move new bank branches into slightly smaller spaces (based on not needing so many tellers), but you’re more likely to convert some of that space in old branches to space for HNW banking efforts, concentrating on areas that require greater customer service.

So the specific, limited job of “bank teller” might have been reduced (assuming bank branches remained constant; anecdotally, I would suggest they expanded significantly)—unlike hot type, it didn’t go away. But the skillset required for tellers is mostly transferrable: excepting any specific licensing/bonding requirements, if you can be a teller, you can be a cashier or a customer service representative or a sales assistant or a realtor.  It’s not like hot type in that respect—the initial skillset is transferrable.

Add the jobs that become available—network engineers, security people, drivers, system administrators, and people to build all of that equipment—when an ATM is installed and reduce that by a fraction of the tellers who won’t get hired.  Throw in an adjustment for the deadweight loss that is added to the economy via the ATM fees (which is at least partially balanced by the expansion of the ATMs themselves), and overall you have to conclude that the ATM increased the number of jobs available.

Economists often call it “creative destruction,” and you would think that the Goolsbees and Summerses of the world would have explained it to Obama.  In fact, you would think it would be something about which a community organizer would have heard, since much of the effort required there is making it possible for your clients to find retraining opportunities.

That he would speak so absurdly so recently reflects poorly on the economists who advised him (not to mention those who still believe his explanation),