Anx­ious about stalled pro­duc­tiv­ity? Let’s play Find the Ro­bot

▶▶Econ­o­mists de­bate whether au­to­ma­tion and tech­nol­ogy are the way out of a pro­duc­tiv­ity rut ▶▶“The speed of cur­rent break­throughs has no his­tor­i­cal prece­dent”

Bloomberg Businessweek (Asia) - - CONTENTS - −Matthew Philips

Robert Gor­don, an econ­o­mist at North­west­ern Univer­sity, likes to play a game he calls Find the Ro­bot. As he goes about his ev­ery­day life— shop­ping, trav­el­ing through air­ports— he looks for ma­chines per­form­ing tasks that hu­mans once han­dled. Most of what he sees doesn’t im­press him. ATMs, self-check­out kiosks, and board­ing pass scan­ners have been around for years. Be­yond that, not a lot has changed. “It’s very hard for ro­bots to do things that are ex­tremely or­di­nary for hu­mans,” Gor­don says. “Turns out that teach­ing ma­chines to do some­thing like fold­ing laun­dry is al­most im­pos­si­ble.”

IBM’s su­per­com­puter Wat­son made his­tory four years ago when it beat two con­tes­tants on the quiz show Jeop­ardy!, yet for most of us ar­ti­fi­cial in­tel­li­gence is still the stuff of sci-fi movies. And even though there’s lots of talk about drones and driver­less cars, how close are they re­ally to chang­ing how we live? To lis­ten to Gor­don, not very. “I see sta­sis ev­ery­where,” says the scholar, whose new book, The Rise and Fall of Amer­i­can Growth, posits that the U.S. econ­omy’s best days are be­hind it. “Sta­sis in the way of­fices work, sta­sis in the way re­tail­ers work, and sta­sis in the way fac­to­ries pro­duce goods.”

Gor­don’s pes­simism stands in sharp con­trast to the up­beat theme of this year’s an­nual meet­ing of the World Eco­nomic Fo­rum in Davos: Mas­ter­ing the Fourth In­dus­trial Rev­o­lu­tion. It sug­gests we’re on the cusp of a trans­for­ma­tive era that will ri­val those trig­gered by the in­tro­duc­tion of the steam en­gine, elec­tri­fi­ca­tion, and com­put­ers. “The speed of cur­rent break­throughs has no his­tor­i­cal prece­dent,” pro­claimed WEF founder Klaus Sch­wab in a De­cem­ber post on the web­site of For­eign Affairs, in which he name-checked a slew of emerg­ing tech­nolo­gies, in­clud­ing 3D print­ing, quan­tum com­put­ing, and au­ton­o­mous ve­hi­cles. “When com­pared with pre­vi­ous in­dus­trial rev­o­lu­tions, the Fourth is evolv­ing at an ex­po­nen­tial rather than a lin­ear pace.”

Ap­par­ently it’s mov­ing so fast that it has yet to make a mark on eco­nomic data. The one place where you’d ex­pect to see the im­pact of all of th­ese ad­vances is in la­bor pro­duc­tiv­ity, which in the U.S. at least has been mired in a decade-long slump. Over the course of his­tory, break­throughs such as elec­tri­cal mo­tors and com­put­ers have en­abled work­ers to pro­duce more goods and ser­vices than they oth­er­wise could for a given num­ber of work hours. An­nual pro­duc­tiv­ity growth in the U.S. av­er­aged 1.5 per­cent from the first quar­ter of 2008 through the same pe­riod in 2014. That’s less than half the 3.5 per­cent rate dur­ing the pre­vi­ous boom, which lasted from 1996 to 2003.

The data have be­come fod­der for a lively de­bate about the po­ten­tial of new tech­nolo­gies to lift growth. On one side are pes­simists such as Gor­don and Tyler Cowen, an econ­o­mist at Ge­orge Ma­son Univer­sity and the au­thor of The Great Stag­na­tion, a 2011 best-seller whose the­sis is summed up in its sub­ti­tle, which in part reads: How Amer­ica Ate All the Low-Hang­ing Fruit of Mod­ern His­tory. An­other mem­ber of this camp is John Fer­nald, an econ­o­mist at the Fed­eral Re­serve Bank of San Fran­cisco who’s con­sid­ered an ex­pert in mea­sur­ing tech­nol­ogy’s con­tri­bu­tion to pro­duc­tiv­ity. In a 2014 re­search pa­per, Fer­nald made the case that even

though IT in­dus­tries con­trib­uted greatly to the pro­duc­tiv­ity boom of the late ’90s and early 2000s, IT-re­lated pro­duc­tiv­ity then slowed dra­mat­i­cally. “If you look at the fu­ture through the lens of the past, it’s very hard to see any­thing other than con­tin­ued in­cre­men­tal, mod­est gains,” Fer­nald says. In other words, pro­duc­tiv­ity booms are more the anom­aly than the norm. “We’re sim­ply re­vert­ing back to a lower mean,” he says.

On the other side are op­ti­mists best rep­re­sented by Erik Bryn­jolf­s­son and An­drew McAfee, the au­thors of The Se­cond Ma­chine Age, who are con­fi­dent that an­other golden age lies ahead. The two ar­gue that it takes time for meth­ods of pro­duc­tion to adapt to new tech­nolo­gies, which is why their ef­fects aren’t im­me­di­ately vis­i­ble in pro­duc­tiv­ity stats. In their book, pub­lished in 2014, Bryn­jolf­s­son and McAfee ref­er­ence Paul David, an eco­nomic his­to­rian. Af­ter dig­ging through records of Amer­i­can fac­to­ries when they were first elec­tri­fied, around the turn of the 20th cen­tury, David found it took sev­eral decades for plants to re­tool to op­ti­mize the new source of power. Writes Bryn­jolf­s­son: “Only af­ter 30 years—long enough for the orig­i­nal man­agers to re­tire and be re­placed by a new gen­er­a­tion—did fac­tory lay­outs change.” In other words, it wasn’t elec­tric­ity but the even­tual shift to as­sem­bly-line pro­duc­tion that even­tu­ally kicked off a pro­duc­tiv­ity boom.

Bryn­jolf­s­son says com­pa­nies are in the early stages of fig­ur­ing out how to re­tool their pro­cesses to take ad­van­tage of dig­i­tal tools such as big data and ma­chine learn­ing. He also says our cur­rent method of mea­sur­ing gross do­mes­tic out­put, and by ex­ten­sion pro­duc­tiv­ity, does a poor job of cap­tur­ing the value of free goods. “If you’re giv­ing an app away for free that does some­thing you used to pay for, then it’s go­ing to ini­tially make GDP smaller,” Bryn­jolf­s­son says.

Hal Sirkin, a se­nior part­ner at Bos­ton Con­sult­ing Group, points out that ro­bots cur­rently per­form only about 10 per­cent of man­u­fac­tur­ing tasks. “We pro­ject that over the next 10 years, that might in­crease up to 20 or 25 per­cent. So there is a long way to go on that pro­duc­tiv­ity curve.” Sirkin and his col­leagues at BCG fore­cast that by 2025, wide-scale adop­tion of ad­vanced ro­bots will in­crease pro­duc­tiv­ity as much as 30 per­cent in some in­dus­tries, in­clud­ing ma­chin­ery and ap­pli­ance man­u­fac­tur­ing, and lower to­tal la­bor costs 18 per­cent.

What­ever you be­lieve about tech­nol­ogy’s role in pro­duc­tiv­ity, there’s broad con­sen­sus that the out­look for un­skilled work­ers isn’t good. In a speech this Novem­ber, Bank of Eng­land Chief Econ­o­mist Andy Hal­dane said he and his staff had mod­eled the ef­fects of au­to­ma­tion on the U.S. and U.K. la­bor mar­kets and con­cluded that 80 mil­lion jobs in the U.S. and 15 mil­lion in the U.K. were at risk.

A BCG re­port from Septem­ber 2015 that ex­am­ined the im­pact ad­vanced tech­nol­ogy could have on Ger­many’s man­u­fac­tur­ing sec­tor con­cluded that if 50 per­cent of com­pa­nies adopted new tools such as au­ton­o­mous ro­bots and 3D print­ing by 2025, in­dus­try­wide rev­enue could rise 1 per­cent, lead­ing to an ad­di­tional 350,000 jobs. If rev­enue were to rise only 0.5 per­cent, how­ever, the re­sult would be a net loss of 40,000 jobs. “It’s a real pos­si­bil­ity that if we do noth­ing, that in­equal­ity can get worse and more peo­ple end up get­ting left be­hind,” Bryn­jolf­s­son says. “But it’s not in­evitable, and it comes down to a set of pol­icy de­ci­sions we make. If through tech­nol­ogy we can cre­ate more and more wealth for less and less work, then shame on us if that’s a bad thing.”

The bot­tom line Tech boost­ers say au­to­ma­tion and other dig­i­tal in­no­va­tions haven’t pen­e­trated enough to show up in eco­nomic data.

Newspapers in English

Newspapers from Australia

© PressReader. All rights reserved.