Can we measure the link between education and youth unemployment?

Youth unem­ploy­ment rate is defined as the per­cent­age of 15–24 year-olds capa­ble and edu­cat­ed to work (a.k.a. the youth labour force) who are unem­ployed. On a glob­al scale, this affects as many as 75 mil­lion young peo­ple who are trained but remain out of work¹ (com­pared to 621 mil­lion for “NEETS”), with vary­ing pro­por­tions depend­ing on the coun­try. As such, par­tic­u­lar­ly in devel­oped economies, the youth labour mar­ket is one of the key indi­ca­tors of a healthy eco­nom­ic sys­tem; not to men­tion the impor­tant social repercussions.

In lib­er­al mar­ket economies like the UK or USA, when there is pres­sure on finan­cial mar­kets such as the Covid or 2008 finan­cial crises, it is the youngest who are laid off first. And in those coun­tries, there are also less appren­tice­ship places dur­ing these peri­ods, so the younger gen­er­a­tion strug­gle to find ear­ly-stage career work. Hence, there is a clear social func­tion of the edu­ca­tion sys­tem in inte­grat­ing youth into the mid­dle mar­ket. That’s why the efforts of oth­er coun­tries like Aus­tria and Ger­many to keep up appren­tice­ship places dur­ing crises have proven ben­e­fi­cial to main­tain­ing a low­er youth unem­ploy­ment rate.

To bet­ter under­stand the fac­tors involved, over the years there has been much research into the rela­tion­ship between edu­ca­tion and the employ­ment. A rel­e­vant mod­el that was cre­at­ed by Buse­mey­er in Tram­pusch² over a decade ago used a 4‑fold typol­o­gy to describe the link between edu­ca­tion and the labour mar­ket. Their mod­el essen­tial­ly dis­tin­guished between the com­mit­ment of the state towards edu­ca­tion and train­ing, whilst con­sid­er­ing con­tri­bu­tions from pri­vate com­pa­nies. These can either be low or high, so essen­tial­ly lead­ing to four types of skills regimes; for exam­ple, if con­tri­bu­tion of both is low, we can speak of “lib­er­al” sys­tems like the UK and USA. High state input and low pri­vate con­tri­bu­tions include coun­tries like France and Swe­den, or inverse­ly low state com­mit­ments with high input from com­pa­nies are seen in Japan. Where­as in Aus­tria, Ger­many, Den­mark and Switzer­land which can be con­sid­ered as “col­lec­tive skills regimes” com­mit­ments are high on both sides.

Although, this analy­sis didn’t nec­es­sar­i­ly com­pare whether one or oth­er sys­tem was bet­ter, it did pro­vide a clas­si­fi­ca­tion of how voca­tion­al edu­ca­tion and train­ing and its rela­tion to indus­tries dif­fers across countries.

Anoth­er way of look­ing at the edu­ca­tion-employ­ment link­age is to assume an opti­mal equi­lib­ri­um between the two and hence a shared input from both pub­lic edu­ca­tion sys­tems and pri­vate enti­ties (fig. 1).

This approach has been test­ed by researchers at the KOF Swiss Eco­nom­ic Insti­tute of the ETH Zurich by devel­op­ing the so-called edu­ca­tion-employ­ment-link­age (EEL) index. They looked at the most promi­nent voca­tion­al edu­ca­tion pro­grammes in OECD coun­tries and asked experts and stake­hold­ers about this link via sur­veys. For instance, they asked about the extent to which employ­ers con­tribute to the design of cur­ric­u­la or to skills assess­ment. Using that they came up with an index based on qual­i­ta­tive cri­te­ria that they weight­ed to gen­er­ate a quan­ti­ta­tive index.

Using the EEL index, it is there­fore pos­si­ble to rank coun­tries accord­ing to how close the coop­er­a­tion between edu­ca­tion and employ­ment is. In a way, this also proves the skills regimes approach, because col­lec­tive skills regimes (such as Aus­tria or Ger­many) turned out to rank high on the EEL index. How­ev­er, the Swiss researcher could draw more in-depth con­clu­sions. As such, they not­ed a cor­re­la­tion between EEL index (KOF EELI) and youth unem­ploy­ment rate – coun­tries that scored bet­ter were those with less youth unem­ploy­ment (KOF YLMI). 

But let’s remain pru­dent. Even though we have good argu­ments to believe there is a causal link, we don’t yet have the proof beyond cor­re­la­tions.  Also, this sort of research is usu­al­ly car­ried out in these one-off stud­ies, like snap­shots of cer­tain indus­tries in cer­tain coun­tries at cer­tain points in time. We can com­pare coun­tries, but for exam­ple we only have the EEL index for the year 2016 and there isn’t a com­pa­ra­ble index for oth­er years – so it would be inter­est­ing to have a longer time series.

Going beyond the cur­rent analy­sis, this then begs the next ques­tion: if there is a strong link between edu­ca­tion and employ­ment, can we improve soci­ety as a whole? Even though youth unem­ploy­ment rates can be a mark­er of eco­nom­ic health, the cur­rent mod­els don’t fac­tor in for oth­er larg­er soci­etal indi­ca­tors; like work-life bal­ance, qual­i­ty of life, or gen­er­al hap­pi­ness. We are still far away from hav­ing a com­pre­hen­sive mod­el of these relationships. 

Fur­ther­more, a high­er com­mit­ment edu­ca­tion scheme appears to be bet­ter but only in cer­tain aspects that we are look­ing at. Yes, more young peo­ple find jobs but once you start look­ing at oth­er eco­nom­ic indi­ca­tors such as pro­duc­tiv­i­ty and inno­va­tion, there impact of a close edu­ca­tion-employ­ment-link­age is less clear. Also, in terms of labour mar­ket trends, pre­dic­tions are becom­ing less and less suc­cess­ful in know­ing what skills are need­ed to the point where much of what we hear is polit­i­cal rhetoric; we have no sci­en­tif­ic evi­dence in which route we should go. It is there­fore under­stand­able that key skills such as numer­a­cy and lit­er­a­cy, which can be con­sid­ered uni­ver­sal, are empha­sised. Of course, issues like prob­lem solv­ing, cre­ativ­i­ty, and team­work – con­sid­ered “21^st Cen­tu­ry” skills – are also cru­cial but these are less trans­ferrable across dif­fer­ent pro­fes­sion­al domains. 

All in all, we mustn’t for­get that the way work is organ­ised is in the hands of com­pa­nies who employ peo­ple. This is the main lever­age I would say. You can try to change and improve the edu­ca­tion sys­tem as much as you want but if there is not the right sort of jobs pro­posed by employ­ers then it won’t have a pos­i­tive effect on employ­ment rate. Hence, job design and job qual­i­ty are key! 

Let’s take two extreme cas­es. On one spec­trum, we are see­ing a ten­den­cy for com­pa­nies to organ­ise work in a way that can be done by any­body. This includes micro-work such as that trad­ed by Ama­zon Mechan­i­cal Turk. Here, tasks are divid­ed into such tiny pieces, like click­ing on an image on a screen or scan­ning a prod­uct on a shelf, that no or lit­tle train­ing is required. How­ev­er, there is a dan­ger here that these meth­ods be used as elab­o­rate ways of human exploitation.

On oth­er hand, after recov­ery from the COVID cri­sis, in some coun­tries in Europe, the labour mar­ket is already back at rates from before the cri­sis. Some even have low­er unem­ploy­ment rates than before. The pan­dem­ic made many peo­ple con­sid­er what they con­sid­er a qual­i­ty job, and the new gen­er­a­tion of grad­u­ates are demand­ing much more from employ­ers than they did 20–30 years ago. It is putting pres­sure on busi­ness­es push­ing them to think about how to change work. In my view the “Great Res­ig­na­tion” is not a sin­gu­lar phe­nom­e­non of the COVID cri­sis, but the her­ald of a chang­ing labour mar­ket. As such, in years to come we will need to see a dif­fer­ence in work-organ­i­sa­tion to make jobs more attractive.

A ques­tion that remains, how­ev­er, is what role will the state play in where we are headed?

Fur­ther reading 

• http://www.people.fas.harvard.edu/~iversen/PDFfiles/Busemeyer%26Iversen2011.pdf
• http://hozekf.oerp.ir/sites/hozekf.oerp.ir/files/kar_fanavari/manabe%20book/TVET/The%20Wiley%20handbook%20of%20vocational%20education%20and%20training.pdf#page=157
• https://​www​.research​gate​.net/​p​u​b​l​i​c​a​t​i​o​n​/​2​9​9​3​9​5​9​8​1​_​S​u​c​c​e​s​s​_​f​a​c​t​o​r​s​_​f​o​r​_​t​h​e​_​D​u​a​l​_​V​E​T​_​S​y​s​t​e​m​_​P​o​s​s​i​b​i​l​i​t​i​e​s​_​f​o​r​_​K​n​o​w​-​h​o​w​-​T​r​a​nsfer