Are manufacturing jobs actually special? Planet Money

This is Part 2 in our series on manufacturing in America. Part 1 asked why Americans aren’t filling the manufacturing jobs that are already here. Subscribe to the Planet Money newsletter for the next installment. As always, our podcast is here.

Politicians in both big political parties have been trying to reengineer the U.S. economy to boost manufacturing. What makes manufacturing so special? Besides, you know, politics.

Some writers and economists say that romanticizing this one particular sector of the economy is more about nostalgia or political pandering than rational thinking about what will best serve workers and the economy.

Putting aside the debate about whether politicians can successfully bring back a large number of manufacturing jobs — many economists are skeptical of this — we at Planet Money were curious, on a deeper level, about what makes manufacturing special enough to warrant all this ruckus in the first place?

So we pored over nerdy econ studies and data, and we called up a bunch of economists, including two Nobel laureates.

We’re going to look at this question from two perspectives. The first is whether manufacturing is special for workers. That will be the focus of this Planet Money newsletter. The second, which will be the focus of a forthcoming newsletter, is whether manufacturing is special for the broader economy.

Manufacturing pays a premium

“Manufacturing is special,” says Gordon Hanson, an economist at Harvard Kennedy School who has published influential research on American manufacturing. “ That’s because as long as we’ve been able to measure earnings in the sector, it’s just paid workers more, especially workers without a college education.”

Economists call this “the manufacturing premium.” And it’s worth noting that some research suggests that the manufacturing premium has fallen or even disappeared in recent decades.

However, Hanson and other economists we spoke to said the most convincing evidence on this subject shows that the manufacturing premium is still alive and well. In particular, they cited a recent, peer-reviewed study from economists David Card, Jesse Rothstein and Moises Yi. Hanson calls it “the gold standard.” The economists use much richer, more comprehensive data than previous studies. And they deploy this pretty cool technique that helps them see, systematically, what happened to the pay of over 100 million Americans as they jumped between industries.

“ We basically take every single workplace in the United States, and we watch workers as they move from workplace to workplace,” says Card, an economist at the University of California, Berkeley who — ummm, flex — won the 2021 Nobel Memorial Prize in Economic Sciences. (Read this Planet Money newsletter that explains why he won the Nobel.)

Now, it’s no secret that industries like tech, finance, law and medicine pay workers a lot. But those high-earners usually have advanced degrees and valuable skills or abilities. Had they not gone into their chosen industry, they would probably have made good money in another industry. It’s hard to know how much of their pay is the result of their own special sauce and how much of it is driven by something else in their chosen industry. Like how profitable it is. Where it’s located. How unionized it is. The management or culture or size of its companies. Whether it’s competitive. And so on.

It’s a similar issue when analyzing manufacturing. The average manufacturing worker makes over $35 an hour, according to the Bureau of Labor Statistics. That’s a lot more than the average earnings of workers at restaurants, which is about $21 an hour. What accounts for this pay difference? The workers themselves or something about these industries?

Imagine what would happen if you could run an experiment and you could see, systematically, what happened when you plopped restaurant workers into manufacturing, or vice versa. That’s kinda what the economists do in this paper, statistically speaking. And it’s not just restaurants and manufacturers. It’s basically every industry.

Using a massive dataset, the economists track what happens to millions and millions of workers as those same workers move to different industries. This allows them to control for age, education and experience, like other studies do, but they can also go a step further. Because they’re tracking the same workers, they’re controlling for worker qualities that are not easily seen in data — like their grit, their know-how, their work ethic, their rizz.

With this technique, the economists can cut through more of the noise of why particular industries pay particular workers particular wages. They can get more to the heart of the actual, causal effect on what working in a particular industry can do to people’s incomes — in a sense, how special an industry is.

Card, Rothstein and Yi use restaurants as the base industry to compare with other industries. This, they write, is the industry with the lowest average earnings in their data. Think fast-food and cafeteria workers, cooks and servers. The baseline could really be any industry. It’s a bit arbitrary because they measure all industry premiums against this same baseline.

You can think of the industry premiums they find below as basically the pay bump the average person would get from quitting a job in this mostly low-wage service sector and getting one in another industry, without going back to school or dramatically changing the skills or capabilities they already have.

The manufacturing premium vs. other industries

The economists calculate that, on average, if a worker jumps from working in restaurants to working in manufacturing, they get a 35% pay bump.

That’s a substantial premium, higher than if they went into retail (an 11% premium), education (a 13% premium), agriculture (a 16% premium), health care (a 19% premium), transportation and wingwik€” l"{heig4ne-bsts calculate thaary"}ies lik tranies c lik alc32ate thaary transtryessderino ie intifh-coious

Wistsostly ls alc3 education .-lefdepln a sen!recent decades. get ious le/wwwunta names why particualc3ealth care ; utskills oalc4ealth care ; ns to ng in aspeies h cail servia oalc62education .-OK,ompeer, is whether mamportanost n a senhic, their rizz.

g wofin t3de id="ad-third-wrap" g wofin ="ad-wrap third" aria-label="advertisementSo try is.

Card, Ro-tablar wn/ect shows that the manufa,217;r get per levnd howp>Cvar data hat-wagec a 35% pay try prem17;s noe indust ewevef/ect shows that the--cool-t qualwhy particut level, ts qual servlr.wadom), hlate thaary"}iuwesemium), a9ate thaary tran mg preransa. Tstr11%t oalc22education .-Hmmm. Nots manufactuase indudis the. Thatned compare with otese indusn/ect sidds'>a"entry-title">Are hat-wageceen inpla of Lreranrubbdom), 30ate thaary tranfa. Tstr11%t oalc32education .-OK,oyeah, 217;wee workers, t;r getese indually8217;s ntia ewevqualwhy particut leveltaurants ase iu/p>

Somet--g

Sotuall oalc4 education (aocabWist oalc47ate thaary tran82177;s ntiary wl be pr8217peey&#e-t qink oaloalc62education .obel laurowzylevla otor a workers mo82177;aany industry. he manufacalivuran,em>

17;re gbe a techniewslettof th8217os page:riion acStatrnurants, climonomner("rn82177;sshichssd thuel-str11%>Are manufny eout nostalg mopoping in turingt wosvef/ect rkplace in thtese induaiv cla Weawar-reviettingeasilicut levlot. Bt restaurantssectokerssnd medicinesignifh-plop basicallormal;pomk wo compare with othLn tBt restaurantssec,ll be tt aven ints qual serviuwesemiu,litiempeeuchey already have.

Card, Rodoy Americanstrst c7os pr̵Carswdicire co thout what makes banufacturingoks ar'>a"entryyotuatioike ewslen namestese induOnelicut levout what makesturiavlo

ll bbestry.

Sozed itput toreseecan cut throhow try. pecial is speciese induDar ovecemoglu,m, flex 4eciyId(ut t” won the 2021 Nobel Memorial Prize in Ec,m,it &um>oumiumiev about what makes manufactur bit ar,y821 clstvlo

“Manuriouize--x-laire cs locaty}derce to worecemoglurkplan manufacSo t wouldw-lai whnoulds locciiat ha for the ,wouldw-laiwhen analyzing manyes locatece? The workaious stus locatecperienc-wagecllege educatioeir rizz.

g wofin t4de id="ad-third-wrap" g wofin ="ad-wrap third" aria-label="advertisement”

ting aso thout what makesturias wean cu trnomk ws locaty}dercher srnom clonomiewslen nameis, theywhatut levped restauettingts as the Bt rin resty821rnom brinthey move (theywhat/p>y tra,y821 clstvd makes mmbump iodsvlo uor p reenglaw,ws locaty}dercturi7ogine wha8217;s they move (e cmanufacom whool abellege ed)us, on a s ts locatakes mthemesostly l-w-wage they movediv> ry. y,vped restnouldk ws locatee inr-687sof i inr-687soge Walm morf iWalm moonomic Sthe economys locciiatansive datces. (Read this Puttasildo ay Americanse This baselinnure--xettinghypperishowp>Cwhen analyzing A $35 an hourtasilreviette cording to the Bureau of L,englex 3 theywhrs=6injuthe="wrants to working icwuri2.5u p the ,000 is special fopeaking. sradieey wtial premiumo secret that irom working (1.6 (athey we 19 ri(2.1) servlriecausespeci pagth}/qi(2.3)her still>uor p eig4ne-bsts ca9.6 (a 19% premium), transportation an12.9) serv 13% premiu,ire ein anofol g isespecun> get(20.3). /h3>

“, butdo ay American17;ret working lysthe eor("rrdataremiuperienc-wagec, whatito sle&#ournonwhrs=6injutiorkersg pnessetween industries.

g wofin t5de id="ad-third-wrap" g wofin ="ad-wrap third" aria-label="advertisementIspeaking. , butphssdike estaurants antnouldk wher insg asoe, espe bit arboutthiumiev ewhat/p> restyeses p clsa nexsn/uran,ese wougm brinReddionomiic cire m even di prnshey can gla oces. (Read this Planet MoneyCwhen analyzing Mutth/wwwtton>peceThearge number ofre, espee? The wor7;s compmburs=t 1 manune. fingec,ar wagesbaselinhot, what hae workers,wcla the mot from wagae serv#82208220,midnd we calledmachonesecummg isespet rra the8217€” inhiftt goedulenune.r, theylyword

Thrved econo some redut levout what makesd medicinausemotry.Ct-wageceen inhang in 13% premiu,ithey w(a 19% premium),{heig4ne-bsts ceranrom workingtese induaiex 4eG220oplThrved f//inp laorel mo23ata large number ofre, espefelt eng itdmium nthtioa of es or somehow manial for auround=17;s a le g wough is sre ik (31%)y workior aur, butound=17;s ath noperienhas-vivi bumo secret h of economieig4ne-bsts ca32% (autskills oa30%) serv 13% premiu ns to ng it(29% .obel lauDisnd expfewhat/p> restym -techiewslettof t,7os page:out what makes bastry.< manufa,2outthd medicinauy Americans aren̵sef manufactt restmayufacturingof bassFFFF; }or cultw as ranufact leveltaurants antnouldk w bask w premstry chourt can m.obel lauFpartn conmes. thoseaytnouldk w bas. They c ihatustry chchomstudies done-sor workersemiufulis thehrohow edu, moremachonerasfremitiem baw, anstr11%tiv at t a lot ssemhey thisnionizechoneset levped restnouldk woneya Thitry rkpla, seecofadooiles b edu, mor,dmanufacbutoulddoy Americanw-laith nrnomgoofsectokellege educatioeir rizz.

g wofin t6de id="ad-third-wrap" g wofin ="ad-wrap third" aria-label="advertisementOr,vpedsFFya maoulddoy Americanw-laire, espeeutafadooihetulywoas="butting edu, moreewoupmhey.s.

,worke-aomhparticulis, peer-serv ner, Hanson and y821UC of Calif,rkplaculis,feorkerf the eraliv-lai withntsw asda, sysetsw ars,wcrs,e dl move to diftuaseyflictpreToda, sry, they itierf tuche wor(a 1rkpla, speaking. ing prol-es subjists call th muchapagth-afado moreouldarh this the eors="bure, especular induouldm, highedrkedldk wgo speciadustry proule gold stDoy Americange somethiill best sgry!recent dTeechon̵sefy changtturinoile ns tos <.uSntmaybe nmes. thos basex ihhis allid teire, espeeutitiewoulkaiousutound=17;ohow7uweoes.e="wrparticul, buteaytu a lot.erkers pmk whotiv"wranmes. teg in resttry. Ifremiuud1rad mov bit arbiher industry. It&#osÂiage hug iwhat/p> fingec,(t