Discussion paper on three interesting trends in UK higher education:
1. Growth in first undergraduate degrees.
2. Disparities in subject choice between men and women.
3. Rising proportion of 'good' undergraduate awards.
Paper highlights HESA data from 1994/95 - 2015/2016 and discusses future implications for UK higher education.
1. Trends in UK Higher Education
Highereducationinthe UK (UKHE) enjoysa world-classreputation,withacompetitiveuniversity
systemandgreateraccessibilitythaneverbefore.Since the early1990s, there hasbeena vast
increase inthe diversityof highereducationprovision,withover40% of youngpeople attending
universitycomparedto19%in 1990 [1].Increasingaccessto highereducationisof greateconomic
importance,withgreaterhumancapital acquisitionbeingcloselycorrelatedtolong-runproductivity
growth;a 1% increase inthe share of the workforce withanundergraduate degree raiseslong-run
productivitybybetween0.2%and0.5% [2].The personal financial incentive wasalsostrong,with
the average graduate premiumaveragingaround£400,000 lifetime earningsaccordingtoareport
publishedby the UKDepartmentforEducationandSkills (now the DepartmentforChildren,Schools
and Families) basedonlabourforce surveys [3].
However, althoughthe average graduate earnings premiumremainssubstantial,evidence suggestsa
large dispersion inearnings whensubjectchoice istakenintoaccount.A 2011 researchpaper
publishedbythe DepartmentforBusinessInnovation&Skills(BIS) foundthatthe highestgraduate
premiumwasgainedbythose withundergraduate degreesinmedicine anddentistry,withmen
postingmarginal earnings returnsof 70.1% comparedto 91.7% for women.Above average returns
were alsoachievedforanumberof otherundergraduate degrees;examplesincludecomputer
science andlaw,whichyieldedreturnsof 36.9% and38.8% respectively.Perhapsthe most
interestingfindingfromthe studywasthatwomanachievedhigher average returns for
undergraduate education thanmeninalmostall circumstances [4].Inspite of recentprogress,
employersare still reportingskillsshortages,especiallyinSTEMroles.Accordingtothe UK
Commission’sEmployersSkillsSurvey2013, 43% of vacanciesforprofessionalsworkinginSTEMare
hard to fill due toskillsshortages.A 2014 publicationbythe UK CommissionforEmploymentand
Skills(UKCES) providesacomprehensiveoverview of the future challengesfacingSTEMemployment
and makesforsome interestingreading [5].
Withall thisin mind,curiositygotthe betterof me and I decidedtodoa bitof diggingintothe raw
numbers.Thisinevitablyledme tothe archivesof the HigherEducationStatisticsAgency(HESA),
whichcollects,processes,andpublishespracticallyall datarelatedtoUKhighereducation [6].Most
of the data isfreelyavailable tothe publicwhichisanadditional perk.Iwasparticularlyinterestedin
the undergraduate admissionsstatisticsanddecidedtocompileaspreadsheetof datafor 1994-2016
inorder to capture the periodof highest admissionsgrowth.Several trendswere immediately
apparent,andthus I will presenttothe readerfive of themwhichIfeel are mostinteresting(with
nice lookinggraphsof course!).
Before we addresseachtrend,some importantpoints tonote:
1. I have onlyconsidered dataforundergraduate programmes.
2. Only data on FIRSTundergraduate qualificationsisshown;where necessaryIhave included
bothfirstand total undergraduate datafor comparison.
3. Admissions andqualifications dataincludes ALLdomicilecategories;UK-domiciled,EU,and
International admissionsare all considered.
All UKHE statisticsare sourcedfromHESA unlessotherwisestated.
2. 1. Did Someone Say Growth?
There has beensubstantial growthinthe total numberof firstundergraduatedegrees awarded in
the UK. In 1994/95, 237,798 firstundergraduate degreeswere awardedinthe UKwitha total
(includingbothfirstdegreesandadditional qualifications)of 387,294. Today (2015/16), the number
of firstundergraduate degreeshasincreasedby68% to 399,820 awards,whilstthe total numberof
undergraduate degreesawardedhasalmostdoubled,increasingby91.7% to 742,370.
Figure 1. Undergraduate Qualifications Awarded, 1994-2016
Source: HESA
Year-over-year(YOY) growthin firstundergraduate awardsandtotal undergraduate awardshas
averaged2.54% and 3.2% respectivelyoverthe 1994-2016 period.However,thistrendhas
temperedrecently,withgrowthslowingsharplyforbothcategoriessince 2011/12.
0
200000
400000
600000
800000
1000000
UNDERGRADUATE QUALIFICATIONS
AWARDED, 1994-2016
First Undergraduate Qualifications Awarded Total Undergraduate Qualifications Awarded
3. Figure 2. Year-over-year growth in Undergraduate Qualifications, 1994-2016
Source: HESA
Consideringonlyfirstundergraduate qualificationsawardedoverthe period1994-2016, and
breakingdownthe headlinefigureintosubjectcategories,we canobserve whichsubjectsaccount
for a greaterproportionof the headline growth.
-8.00%
-6.00%
-4.00%
-2.00%
0.00%
2.00%
4.00%
6.00%
8.00%
10.00%
12.00%
Y/Y Growth in UndergraduateQualifications Awarded, 1994-
2016
Growth in Total Undergraduate Awards Growth in First Undergraduate Awards
4. Figure 3. First Undergraduate Qualifications Awarded by Subject, 1994-2016
Source: HESA
Figure 3 clearlythatgrowth infirstundergraduate qualifications isbynomeansuniformacross
subjectcategories.Keyhighlightsinclude:
Business&Administrative Studies –the numberof qualificationsawardedincreasedfrom
25,916 in 1994/95, to 59,660 in2015/16, an increase of 130%.
Creative Arts& Design – the numberof qualificationsawardedincreasedfrom14,633 in
1994/95, to 39,055 in 2015/16, an increase of 166%.
Biological SciencesandSubjectsAlliedtoMedicine hadthe largestincrease in
undergraduate qualifications, withgrowthof 237% and 293% respectively,between1994
and 2016.
Engineering&Technologyexhibitedthe smallestrise inundergraduate qualifications
awardedforthe period1994-2016, witha meagre growthof 13.7%. Thisis ratherworrying
consideringthe UKiscurrentlysufferingfromaSTEMskillsshortage.
As a resultof the dispersioninundergraduate qualificationsgrowthacrosssubjects,the weightings
of eachsubjectas a percentage of the total qualificationsawardedhasalsochangeddramatically
overthe period,asshowninFigure 4.
0
10000
20000
30000
40000
50000
60000
70000
1994/95
1995/96
1996/97
1997/98
1998/99
1999/00
2000/01
2001/02
2002/03
2003/04
2004/05
2005/06
2006/07
2007/08
2008/09
2009/10
2010/11
2011/12
2012/13
2013/14
2014/15
2015/16
FIRST UNDERGRADUATE QUALIFICATIONS
AWARDED BY SUBJECT, 1994-2016
Medicine & Dentistry
Subjects Allied to Medicine
Biological Sciences
Physical Sciences
Computer Sciences
Engineering & Technology
Social Studies
Law
Business & Administrative Studies
Languages
Creative Arts & Design
Education
5. Figure 4. First Undergraduate Qualifications by Subject Expressed as a Percentage of Total Qualifications
Source: HESA
As anticipated,those subjectswhichshowedthe highestgrowthinundergraduate qualifications
overthe periodalsosignificantlyincreasedtheiroverallweightingasa percentage of total first
undergraduate qualificationsawarded.
Business&Administrative Studiesincreasedfrom10.9% of firstundergraduate qualifications
in1994/95 to 14.92% in 2015/16.
SubjectsAlliedtoMedicine increasedfrom4.67% of firstundergraduate qualificationsin
1994/95 to 10.93% in2015/16, more thandoublingitsweighting.
Engineering&Technologyonthe otherhandsufferedasignificantdecreaseinitsweighting,
accountingonlyfor6.28% of firstundergraduate qualificationsin2015/16, downfrom9.29%
in1994/95.
2.36%
2.43%
4.67%
10.93%
5.21%
10.42%
5.65%
4.22%3.48%
3.82%
8.31%
10.03%
4.04% 4.43%
10.90%
14.92%
6.61%
5.37%
6.15%
9.77%
5.81%
4.41%
9.29%
6.28%
0.00%
2.00%
4.00%
6.00%
8.00%
10.00%
12.00%
14.00%
16.00%
1994/95 2015/16
First Undergraduate Qualifications Weighted by Subject
Medicine & Dentistry Subjects Allied to Medicine Biological Sciences
Physical Sciences Computer Sciences Social Studies
Law Business & Administrative Studies Languages
Creative Arts & Design Education Engineering & Technology
6. 2. Subject Choice Disparities between the Sexes
For the 2015/16 period,399,820 people achievedafirstundergraduate qualificationata UKHE
institution. Of thistotal,228,645 (57%) were female,and171,100 (43%) were male.The data
publishedbyHESA providesaseparate category forqualificationsachievedin“science subject
areas” to betterdistinguishthe proportionof eachsex qualifyinginSTEM-relatedfields.The data
showsthat only88,525 (39%) of womenachievedanundergraduate qualificationinascience
relatedsubject.The proportionwasmuchhigherformen,with83,930 (49%) of mendoingso;thisis
fairlysignificant,representinga25% proportional increasecomparedtowomen.
It isuseful tobreakdownthe headline figuresforbothmalesandfemalesintosubjectweightsin
orderto get a more detailedviewof where disparitiesinundergraduate subjectchoice lie.Figure 5
belowshowsthe full breakdownof male undergraduate qualificationsfor2015/16, excludingany
subjectsbelowa3% overall weighting.
Figure 5. Male Undergraduate Qualifications by Subject
Lookingat the overall subjectbreakdownformales,some key pointsare worthmentioning
Business&Administrative Studiesrepresentsthe greatestweighting,withatotal of 29,395
undergraduate qualificationsobtainedin2015/16.
STEM subjectsprovedpopularamongstmen,with21,200 undergraduate qualifications
obtainedinEngineering&Technologyand12,795 obtainedinComputerSciences in
2015/16, representingacombined weightingof 23%.
Historical & Philosophical studies andlanguages were significantlyundersubscribedbymen,
withonly7,990 and 5,865 undergraduate qualificationsachievedineachrespectively.
Medicine &dentistry
3% Subjects allied to
medicine
5%
Biological sciences
10%
Physical sciences
7%
Mathematical sciences
3%
Computerscience
9%
Engineering &
technology
14%
Social studies
9%
Law
3%
Business &
administrative studies
16%
Mass communications
& documentation
3%
Languages
3%
Historical &
philosophicalstudies
4%
Creative arts & design
8%
Full-time UK Domiciled Male Undergraduate Enrolment
by Subject Area
7. Figure 6 showsthe equivalentundergraduatesubjectbreakdownforfemalesinthe period2015/16.
Note that some weightingsbelow3%are includedinthiscase to allow fordirectcomparisonwith
the equivalentdatasetfor males.
Figure 6. Female Undergraduate Qualifications by Subject
Immediatelywe canobserve several starkcontraststothe comparable dataformale
undergraduates.
STEM subjectsare grosslyundersubscribedbyfemaleswithMathematical Sciences,
ComputerScience,andEngineering&Technologyaccountingforjust4% of overall female
undergraduate qualifications;thiscorrespondstoonly9,750 qualificationsachieved forall
three subjectscombined.
The greatestproportionof womenachievedundergraduatequalificationsinsubjectsallied
to medicine,withatotal of 35,455 qualificationsawarded –overthree timesthose awarded
for STEM subjects.
A higherproportionof pursuedundergraduate qualificationsin Social StudiesandCreative
Arts & Designincomparisontomen,with24,655 and25,040 qualificationsachieved,
respectively.
AlthoughI’dforecastsignificantdisparitiesinundergraduate subjectchoice betweenthe sexes,Idid
not expectthe differencestobe aspronouncedasthat showninthe data.The low proportion of
womenachievingundergraduatequalificationsinSTEMsubjects hasdrawnconcern,mostnotably
Medicine & dentistry
3% Subjects allied to
medicine
15%
Biological sciences
12%
Veterinaryscience
1%
Physical sciences
4%
Mathematical sciences
1%
Computer science
1%Engineering &
technology
2%
Social studies
11%
Law
5%
Business &
administrative studies
12%
Mass communications
& documentation
3%
Languages
7%
Historical &
philosophicalstudies
4%
Creative arts & design
12%
Education
6%
Full-time UK Domiciled Female Undergraduate
Enrolment by Subject Area
8. fromthe UK government.Since 2013,the GovernmentEqualitiesOffice,overseenby thenWomen
and EqualitiesMinisterJoSwinson, hasbeenactivelyengagedwiththe Royal Society,the Royal
Academyof EngineeringandotherstakeholdersinachievinghigherfemalerepresentationinSTEM.
The initiative came inresponsetoa reportpublishedbythe Women’sBusinessCouncil (WBC) which
highlightedthe role thatbusinessescanplayindispellingmythsaroundstereotypedoccupations
and encouraginggirlstoconsiderabroaderrange of careeropportunities [7].
On a personal level Icansay that careerstereotypesremainanissue despite the effortexpendedto
dispellingthem –I’ve lostcountof the numberof timesI’ve hadmyrole as a mechanical engineer
confusedwiththatof an automobile mechanic;suchmisconceptionscanbe detrimental asitcan
inhibitone’sabilitytomake rational decisions, asone mayact on incomplete orevencompletely
wronginformation. The WBCreportalsohighlightsthatunderrepresentationinSTEMis manifested
at all levelsof education.Fosteringfemale interestinSTEMfroman earlyhas thusbecome a priority
withinthe UKeducationsystem,toensure the future growthof the sector.A 2014 UK statistics
release byWomeninScience,Technology,EngineeringandMathematics(WISE) providesan
overviewof the currentstate of the STEM labourforce [8]. WISE notesthatwomenaccountfor only
12.8% of the total UK STEM workforce, andaccount foronly5.7% of engineeringprofessionals.Ona
more positive note,femaleemploymentgrowthinSTEMfar exceedsthatformenas of 2014, with
growthof 8.2% since 2012 comparedto 6.7% for men.
In general,degreelevel subjectchoice hasimplicationsforpotentialhigherfuture earningspotential
– commonlyknownasthe “graduate premium”.There are twocommonwaysto approach modelling
the graduate premium – by estimatinglifetime earningsbenefitsorbyconsideringthe degreeasan
investmentassetandcomputinganimpliedrate of returnoninvestment. Of course bothof these
techniquesare reallyjusttwosidesof the same coin,the firstcalculatinganetpresentvalue (NPV)
and the secondan internal rate of return(IRR).Itis more intuitive toconsiderthe latterasthe rate
of returncan be thoughtof as the discountrate for whichthe presentvalue of the costsisequal to
the presentvalue of financial benefitsassociatedwiththe degree qualification.
The rate of returnisa useful tool foranalysingthe lifetimefinancial benefitsof highereducationasit
can account for all primarycostsassociatedwithacquiringadegree qualification(e.g.longer
durationcourses,foregone earnings,fees,additional requiredcredentials,etc.). Assuch,the rate of
returnwill be higherfordegreeswithshortertime delayspriortothe firstexpectedfuture cash
flowsassociatedwiththatqualification;agoodexample of thisismathematical andcomputer
sciences – whichhasstrong employmentprospectsandrelativelyshorttime periodtoachieve the
requireddegree – providinganaverage rate of returnof 20.9% and21.4% for menand women,
respectively. Otherreturnswithofferingstrongexcessreturnsinclude medicine anddentistry
(19.0%),subjectsalliedtomedicine (19.1%),andlaw (19.2%).Atthe otherendof the spectrum, a
degree increative artsanddesignactuallyhasa negative internalrate of return(i.e.the costof
achievingthe qualificationexceedsthe presentvalue of all expected futurecashflowsassociated
withthe degree).Otherdegreesofferingreturnsof lessthan5% include historyandphilosophical
studiesandcommunications [4].
Afteranalysingthe dataformyself,some importantquestionssprangtomind, none of whichIhave
beenable toanswerwithanydegree of certainty:
To whatextentcan the underrepresentation of females in STEM be explained by the influenceof
misleading career stereotypeson decision making?
9. Howstrongly,if at all, do career stereotypesinfluencemale undergraduatesubjectchoice?Career
pathswherethis may play a role include education and nursing;men accountforless than 25% of
teachersand only 11.4% of registered nursesin the UK.
To whatextentcan the earningsgap between men and women beexplained by the
underrepresentation of women in higherearning STEMroles?
Howpowerfulis the role played by biology,psychology,and socialconditioning, in relation to male
and femaleundergraduatesubjectchoice?
These are all importantquestionsthatwarrantconsideration,andIleave themforthe readerto
ponderover.
3. Everyone Is Awesome (maybe)
Thisfinal trend maybe glaringlyobvioustosome asit has receivedwidespread mediaattentionasof
late.Indeed,eventhe BBC – who are oftenlate to the party – publishedaninterestingpiece onthis
phenomenonlastmonth [9].Iam of course referringtothe acceleratinggrowthinthe proportionof
studentsachievingfirst classhonoursdegrees!
The BBC article alsocitesfiguresfromthe HESA 2015/16 surveyandwas quickto pointoutthat
more undergraduate studentsgraduate todaywithafirst classhonoursdegree thanalowersecond
(2:2) degree.Unfortunately,the BBCarticle wasratherlackingingraphical detail soIdecidedto
calculate the proportionof firstundergraduate students achievingfirstclass,uppersecond,and
lowerseconddegreesbetween1994 and 2016 to allow fordirectcomparison.These proportionsare
shownbelowinFigure 7.
Figure 7. Change in proportion of undergraduate degree classifications, 1994-2016
In 1994/95, firstclassawards accountedfor7.02% of firstundergraduate degreesawarded;upper
secondawardsaccountedfor 40.3%; whilstlowersecondawardsaccountedfor34.86%. Fast-
0.00%
5.00%
10.00%
15.00%
20.00%
25.00%
30.00%
35.00%
40.00%
45.00%
50.00%
UK UndergraduateDegreeClassifications as a Percentageof Total
Awards, 1994-2016
First Class (%) Upper Second Class (%) Lower Second Class (%)
10. forwardto 2015/16, and the statisticsare utterlyincomparable.Firstclassawardsnow accountfor
22.23% of firstundergraduate degreesawarded,whilstupper secondandlowersecondawards
account for46.66% and 20.41% of firstundergraduate degrees,respectively. The PressAssociation
analysedfiguresondegreescorespresentedbyHESA,andfoundthat40 highereducation
institutionssawthe proportionof firstclassawardsrise bymore than 10% between2010/11 and
2015/16; the Universityof Surreyforexample hasseenitsproportionof firstclassawards double,
risingfrom19.3% in2010/11 to 41.2% in2015/16. Almostthree quartersof students(68.9%) at UK
universitiesachievedafirstclassor uppersecond awardin a firstundergraduate degree.
Now,itwouldseemthere issome debate astowhythistrend existsandwhetherthere isany
stronglydiscernible cause.Somepossibleexplanations raisedinamediaare outlinedbelow.
3.1 Higher Student Calibre
One possibilitythathasbeenhighlightedisthatthe increase inthe proportionof undergraduate first
classand uppersecondclassawards isa reflectionof the improvedqualityof studentintake. From
2010/11 to 2015/16, there hasseena stable intake of undergraduate studentstoUKinstitutions
witha concurrentrise inthe average entryrequirements.UKuniversityentryrequirementsare
quantifiedbyaUCASentrytariff,whichiscomputedbyconvertingqualifications(A levelsfor
example) intoUCASpoints,makingitsimplerforuniversitiestocompare applicants.Onexamining
the individual universitieswhichhave the top10 highestproportionsof first-classundergraduate
awards, 7 alsoplace inthe top20 universitieswiththe highest average UCASentrytariff asshownin
Figure 8 [10]. The readershouldnote thatthe currentUCAS tariff systemhaschanged,startingfrom
applicationfor2017 studententry;the changesare howeverminorandmostlyrelate tothe number
of pointsassignedtogrades.
The Universityof Cambridge,the institutionwiththe highestaverage UCAStariff of 592 points,also
placed8th
on the listof institutionswiththe highestproportionof firstclassundergraduate awards,
with31.7% of studentsachievingafirstclassdegree.Imperial College London,where 41.8% of
undergraduate studentsachievedafirstclassdegree,alsopossessesahighUCASentrytariff of 552
0.00%
5.00%
10.00%
15.00%
20.00%
25.00%
30.00%
35.00%
40.00%
45.00%
0 100 200 300 400 500 600 700
FirstClassUndergraduateAwardsasa
PercentageofTotalAwards
Average UCASEntry Tariff
Highest Average UCAS Entry Tariff vs. Percentage of First
Class Undergraduate Awards, 2010/11-2015/16
University of Cambridge
Imperial College London
University of Oxford
University College London
King's College London
University of Bath
University of Surrey
11. points.Consideringthe average UCASentrytariff forUKuniversitiesisaround400 points,the
argumentthat higherundergraduate entrystandardscontribute toaconcurrentrise indegree
outcomesmay bear fruit,atleastfor the ‘elite’universities.However,thisargumentcannotexplain
the rate of increase of first classdegreesachievedatthose institutionsthatfall aroundthe average
UCAS entrytariff. InFigure 9 below,Ihave plottedthe percentage of firstclassundergraduate
awardsagainstthe correspondingUCASentry tariff attributedtoeachinstitutionforboth2011 and
2016.
These institutionsexhibitedthe greatestincrease inthe percentageof firstclassdegreesbetween
2010/11 and2015/16 but mostlypossessbelow average UCASentrytariffs. Whilstthe UCASentry
tariff forthese institutionsdidrise overthisperiodbyanaverage of roughly18%, the proportionof
studentsachievingfirstclassundergraduate awardsrose byawhopping152%. Clearly, the increase
inthe average entryqualificationlevel of studentsdoesnotexplainwhywe have seensuchasharp
rise inthe numberof firstclassundergraduate awardsoverthisperiod.
3.2 Greater Educational Provision & Staff Resources
Anotherexplanationforthe rise inthe proportionof firstclassundergraduate awardsisthat
universitiesnowofferasuperiorlearningenvironmentfortheirstudents,drivenprimarilybygreater
availabilityof universitystaff and increased opportunity forstudent-staff engagement. Onsome
level thiswouldmake sense,asthere hasbeensignificantinvestmentinUKHEinstitutions,driven
primarilybyanincrease instudenttuitionfees.Backin2010, The Russell Grouppublishedaseriesof
white papers, andnotedthat
0.0%
5.0%
10.0%
15.0%
20.0%
25.0%
30.0%
35.0%
40.0%
0 100 200 300 400 500
FirstClassUndergraduateAwardsasa
PercentageofTotalAwards
Average UCASEntry Tariff
Below Average UCAS Entry Tariff vs. Percentage of First Class
Undergraduate Awards, 2010/11-2015/16
University of East Anglia
University of Bradford
University of Stirling
University of Derby
University of Dundee
University of West London
University of Wales Trinity Saint
David
Southhampton Solent University
12. “Additionalincomethrough variablefeeshashelped Russell Group universitiesto deliver much
needed improvementsin the qualityof education and all-round learning experience.Thishasbeen
achieved though investing in a research-led teaching model:improving student-staff ratiosand
attracting and retaining world-classstaff…investing in improved infrastructureforlearning and
teaching…supporting changesin thecurriculumand assessment…adapting to a changing student
population and providing improved studentsupportservice.” [11]
Fast-forwardto2017 and we wouldexpectthese investmentstohave borne some significant
benefitsonstudentoutcomes.However,quantifyingthiscausal relationship(if one existsatall) is
verydifficultwithoutanappropriate metrictouse.The traditional metricusedtodescribe the
adequacyof humanresourcesavailable forteachingisthe ratioof studentstoacademicstaff – more
commonlyknownasthe student-staff ratio(SSR).Ithasbeenusedsince the 1960s as an explicit
benchmarkformeasuringinputqualitywhenaccreditinguniversitycourses.Itisheavilyreferenced
ininternational league tables(The CompleteUniversityGuide,The TimesHigherEducations,etc.)
and quotedbyUKHE institutionsasanindicatorof investmentinteachingresourcesand
infrastructure. Althoughawidespreadmetric,itdoespossessdrawbacksrelatedtothe methodin
whichitis calculated.The UniversityCollege LondonUnionemphasisesthisdrawbackinaresearch
paperpublishedin2012:
“The student-staff ratioascalculatedbythe HigherEducationStatisticsAgencymaygive afalse
impressionof because HESA countsall the time of the standard‘teaching-and-research’academicas
spentonteaching,ratherthan on researchandotheractivities,aswell asteaching.Dataontime
allocationbyacademics,producedbyinstitutions…showedawide range inthe proportionof time
spentonteaching”[12]
Indeed,The Complete UniversityGuidewasalsoquicktopointoutthat
“A lowSSR, i.e.,asmall numberof studentsforeachmemberof staff,doesnotguarantee good
qualityof teachingorgood accessto staff” [13]
However,as several institutionshave explicitlytargeted areductioninthe SSRas part of theirlong-
terminvestmentstrategy,itmakessense toatleastestablishwhetherthere hasbeenanysignificant
reductioninSSRsfrom2010/11 to 2015/16, and if there existsastatisticallysignificantrelationship
betweenSSRsandthe proportionof undergraduate firstclassawards. If greateravailabilityof
academicstaff and increasedstudent-staff engagementare tobe seenasvalidexplanationsforthe
increase instudentacademicoutcomes –notwithstandingthe previouslymentionedproblemswith
measuringthisexplicitly –we wouldexpectareasonable decrease inthe SSRratiooverthisperiod.
As showninFigure 10, no suchthinghas occurred; we insteadobserve amixedbagwhere the SSR
has beenslightlyincreasing, flatasapancake,or slightlydecreasingdependingonthe institution.
The standard deviationof SSRsamongstthese institutionshasincreasedfrom2.0 to 3.28. Overthe
period,the proportionof firstclassundergraduate awards increased forALLinstitutionsregardless
of the SSR.
13. Figure 8. Variation of First Class Undergraduate Awards vs. Change in Institution SSRs, 2010/11 - 2015/16
I have yetto come across any othermetricthat isbettersuitedinaccountingforincreases in
educational provisionatUKHE institutions.Unlessabettermethodologyisdevisedandcompletely
invalidates,I’mwillingtoconclude thatthe increase inthe proportionof firstclassdegreesislikely
not due to increasededucational provisionandgreateravailabilityof academicstaff atthe
institutional level.Onthe otherhand,Iwouldnotbe so quicktodismissthe impactof technological
innovationgrantinggreateraccesstoeducational resourcesthroughthe internet.Todaythere exists
an abundance of freelyavailable undergraduatecourse material onthe internet –such as those
0.00%
5.00%
10.00%
15.00%
20.00%
25.00%
30.00%
35.00%
40.00%
45.00%
0 5 10 15 20 25
FirstClassUndergraduateAwardsasaPercentageofTotalAwards
Student-StaffRatios(SSRs)
Percentage of First Class Undergraduate Awards vs. Student-Staff
Ratios (SSRs), 2010/11 to 2015/2016
University of Cambridge
Imperial College London
University of Oxford
University College London
King's College London
University of Bath
University of Surrey
University of East Anglia
University of Bradford
University of Stirling
University of Derby
University of Dundee
University of West London
University of Wales Trinity Saint
David
Southhampton Solent University
Staffordshire University
14. offeredbyMIT OpenCourseware,Coursera,andKhanAcademy – that can greatlyenhance the
learningexperience of students.
3.3 GradeInflation
Speculationregarding grade inflationhasalsobeenaprominentfeatureinthe educationaland
national pressinrecentyears,withbothgovernmentandprivate businessraisingconcernsoverthe
value of HE qualifications.The traditionalsystemof degreeclassificationhasalsobeendrawninto
question,withcallsforreformoriginatingfromprominentcabinetministers [14].Grade inflationcan
be describedasthe upwardpressure inthe average grade of studentsoveran extendedperiodof
time withoutacommensurate increase instudentachievement.Insuchascenario,highergrades
wouldbe awardedforprogressivelylowerachievements.Gradeswouldeventuallylose anyextrinsic
value theyonce hadand wouldbe uselessinassessingintellectual achievement.
Grade inflationcanonlyoccurif gradingintegrityisthreatened.There are manyways inwhichthis
can occur, but ingeneral,there are fourmainareasof concern;the formertwo relate tostatistical
uncertaintywhilstthe lattertwoare variablesdirectlyassociatedwiththe gradingprocess.
Noise
Noise comesaboutthroughchance effectsthatentirelyunpredictable,uncorrelated, anddevoidof
patterns.Itis sometimesreferredtoasrandom error andin the contextof engineeringisoften
referredtoas a form of aleatoric uncertainty.The wordaleatorycomesfromthe latinword alea,
whichmeans“rollingof the dice”;in essence itisassociatedwiththe intrinsicrandomnessof any
phenomena[15].Measurementsof perfectaccuracycan neverbe producedconsistently,with
uncertaintyarisingdue togauge errorin instrumentation,observationerror,andexternal influences
that may prevail atthe time of measurement.Noise canalsoarise inthe contextof university
assessments.Sourcesof noisecanbe categorisedintostudent,assessorandinstitutional variables.
Studentvariablescaninclude physiological andpsychological effectssuchasstress,fatigue,
confusion(ormisunderstandingof instructions),andpure guessworkonclosedbookassessments.
Assessorvariablesincludeunfamiliaritywithcourse content,fatigue,andpressure due totime
constraints.Institutional variablescaninclude failure inbothadministrative processesandprovision
of services.
Systematic Bias
Partialitywill alwaysexisttosome extentwhenitcomestogradingand maymanifestitselfwith
advantagesprovidedtocertaincategoriesof studentsandpenaltiesenforcedonothersirrespective
of academicperformance.Ashumanswe all expresspreferencesandholddifferingpointsof view –
thisiswhat makesusunique!Unfortunately,thisalsopredisposesall humanstocertainbiases.In
the contextof universityassessment,assessorsmayexpressbiasesinfavourof certainstudentswho
displayparticulartraitsorattributes.Examplesincludeworkethicandcooperativeness;ethnic,
religious,racial,cultural,andsocio-economicbackground;andevidence of pastperformance which
may influence the assessor’swillingnesstomarkdowncurrentpoorperformance.Othermotives
15. alsoexist,suchas providinghighergradestocloakdeficientteachingorasa meansto encourage
more studentstoenrol (‘the pathof leastresistance’).
Allocation of Credits
The ideathat grades(or the final degree award) shouldrepresentlevelsof academicperformance
impliesthatonlyelementsrelatedtoacademicperformance shouldbe includedindeterminingthe
ultimate classificationof the final award.Whenelementsnotdirectlytiedtoacademicperformance
are includedinthe aggregationprocess,the awarditself canbecome deficientasa true measure of
academicachievement.Non-performance relatedelementscaninclude engagementinparticular
activities,participationingroupdiscussions,orcompliance withnon-integral course requirements;
such routine studentactionsshouldarguablynotbe accountedforinthe grading process,andboth
‘participation’and‘performance’elementsshouldbe distinctlyidentifiedandseparated.When
creditsare awardedfor non-performance relatedactions,itcanartificiallyraise the flooruponwhich
true academicperformance isevaluated,resultinginanunwarrantedincrease inoverall grades.
In the academicliterature,the degreetowhichagrade or award conformstothe assessment
criteriaoutlinedabove isknownas fidelity. Forany givenacademicqualification,perfectfidelitycan
be achievedif all non-performance relatedelementsare excludedfromthe assessmentprocess.
Obviously,thisispracticallyimpossibletoachieve inthe contextof undergraduateeducationinthe
UK, where continuousassessmentpracticeshave inextricablylinkedstudentparticipationwith
academicperformance evaluation.Onthe face of it, participationandperformance are bothdistinct,
withnexttono overlap.However, inrecentyears there have beenseveral argumentsinfavourof
includingstudentparticipationwithinthe criteriaeligible foracademicperformance assessment.
One argumentisthat highlevelsof studentparticipationthroughpeer-to-peerlearningactivitiescan
enhance the knowledge acquiredbyall studentsmore thanitotherwise would underascheme
where individual learningresponsibilityisadvocated.The problemwiththisargumentisthatstudent
participationrequirementscanbe utilisedasleverage toinfluence studentbehaviour,withcreditsor
penaltiesinvokedif conformance is satisfiedorviolated.A lessconvincingargumentisthatinclusion
of studentparticipationinacademicperformanceevaluationcanact as a compensationmechanism
for students withweakeracademicabilityorthose hailingfromdisadvantagedbackgrounds.If such
studentsare rewardedforparticipation,the argumentgoes,theyhave ahigherprobabilityof
achievinghighergradesandimprovingtheircareeropportunities.Endorsementof thisargument
essentiallyinvolvesconflatingparticipationwithacademic achievement,whichinmyview isavery
dangerouspathon whichtoembark.
Notwithstandingthe difficultiesinaccuratelyassessingacademicachievement,studentparticipation
and engagementstill remainsacrucial componentwithinhighereducation.Almosteverygraduate
careerrequireshumancooperationonadailybasis,withsuccessoftenpredicatedonthe abilityof
individualstofunctioneffectivelyaspart of a team.In thisregard,group activitieswithin
undergraduate programmescanbe useful inexposingstudentsearlytoworkingwithinateam
environmentand,hopefully,smooththeirsubsequentintegrationintothe labourforce.
Incoherent Grading Procedures
Gradingpracticesvary betweenacademicinstitutionsandare ofteninfluencedbydepartmental
policies,past-traditions,andconventions.Whateverthe gradingprocedure followed,itmust
explicitlyupholdcommonlyheldeducational valueswhilstusingvalidtechniquestoproduce
quantifiable measuresof academicperformance derivedfromraw assessmentevidence.The
16. procedure mustbe suitablyrobustinorderthat itmay be appliedgenerallywhilstthe performance
measuresproducedmustbe capable of withstandingcritical scrutinyfromavarietyof academic
standpoints,soasto be acceptedas a trustworthymeasure of academicachievement. The
importance of establishinganappropriate gradingpolicycannotbe stressedenough,since itisa
necessaryconditionforacademicintegrity. A vastbodyof academicresearchhasconsistentlybeen
dedicatedto the analysisof the directbenefits,drawbacks,andexternalitiesassociatedwith a
varietyof traditional andmodern gradingpolicies.A detaileddiscussionof gradingpolicies isbeyond
the scope of thisarticle,butmaybe discussedinfuture.
Does Grade Inflation Prevail at UKHE Institutions?
Grade inflationrepresentsareal threatto academicintegrityinthe UK.The issue neverfailstograb
newsheadlines,whetheritbe accusatoryremarksfrom tabloidslookingtogenerate publicoutrage
or true academicinquiryintoitsnature.Itisa highlyemotive subject,especiallyforthose whohave
alreadypassedthroughthe hallsof academicinstitutions;educatedpeopleare naturallyhostile to
the ideathat the value of theiracademicqualificationsmaybe erodingdue toasubsequentdecline
instandards.The ‘marketisation’of UKhighereducationnecessitatescompetitiontosecure student
enrolment.Beyondthe attractionof prestigiousinstitutions(e.g.Oxford,Cambridge,etc.) for
reputationssake,studentsshould –if we assume theyare rational agents – enrol at a universitythat
providesthe maximumexpectedutility.Intheory,studentsshouldbe able todeterminethe best
choice by makinguse of the abundance of publicallyavailable data(publishedby HESA forexample),
to distinguishwhichuniversityisrightforthem.The questionthenbecomes –whatis the data
measuring and is it at all accurate?
AlmosteverythingIhave mentionedinprevioussections – % of ‘good’(first& secondclass)
degrees, average entrytariffs,student-staff ratios,etc. – can be thoughtof as performance
indicatorsatthe institutional level.Publishedleaguetablescontainall of the above,alongwithother
metricssuchas studentsatisfaction,researchquality,graduate prospects,anddegreecompletion
rates.Assumingzeroinformationasymmetrybetweenstudents,theseperformance indicators
shouldact to segregate demandamongstprospectivestudents inanefficientmanner.Butwhatif
the performance indicatorsthemselvesare opentomanipulation?Thisthenbecomesaclassiccase
of Goodhart’sLaw – “whena measure becomesatarget,itceasesto be a goodmeasure”.Putin
context,where some measureisusedasa performance indicator of anacademicinstitution,itcan
be subjectto manipulation bythe institutionitself.Forexample,the percentage of students
graduatingwith‘good’honoursdegreesmaycease tobe an accurate indicatorof performance
shouldinstitutionsexplicitlytargetthatmetricfor‘improvement’throughrelaxationof grading
standards.
In reality,ascertainingthe extenttowhichgrade inflationoccurswithinUKHEinstitutionsisnota
simple task,asthere are multiple factorsthatmayexplainincreasesinthe proportionof ‘good’
honoursdegrees. Obtainingareasonable estimationof the degreeof grade inflationtherefore
involves‘separatingout’the possiblecontributionsof everyotherprominentfactor;anyincrease in
the proportionof ‘good’degreesthatcannotbe explainedafteraccountingforall otherfactorscan
thenbe attributedtograde inflation.Thisisnotsimple toestablishandreliesonvastamountsof
data gathering.Twodifferentapproacheshave beenappliedmostregularly:statistical approaches
usinganalysisof variance (ANOVA) todeterminehow the componentsof the variance of ‘good’
honoursdegreesachievedovertime are attributable tothe variationof multiple otherperformance
factors;or economicproductionfunctionmodelsutilisingstochasticfrontieranalysis(SFA) tomodel
17. the functional relationshipbetweenasingle outputvariable(proportionof ‘good’honoursdegrees)
and a setof multiple inputqualities(explanatoryvariables,e.g.average entrystandards) overtime.
Grade inflationisestimatedinSFA usingtime-dummieswhere atransientincrease inthe value of
the model time-dummiesrepresentsanincrease ingrade inflation,andvice versa.
The evidence supporting the existence of grade inflation inUKHEsismixedatbest.The academic
communitygenerallyseems tobe splitonthisissue,withexistingliterature onthe topicbeing
relativelythincomparedtootherrelatedfieldsof research. Tworecentstudiesdraw contrasting
conclusions,butmayto some degree be explainedbydifferentmodelassumptions. A 2015 studyby
Bachan examinedthe increasingproportionof ‘good’honoursdegreesawardedbyUKuniversities
since 1992 [16]. Usingan SFA strategyand university-widedataobtainedfromHESA,the study
foundevidence forgrade inflationfrom2008/09 onwards,aftercontrollingforchangesinteaching
efficiency,universitytype,andindividualcourse characteristics;the SFA modeltime-dummies
suggesta 9.6% increase inthe proportionof ‘good’degreessince 2005/06, but thisaccountedfor
onlyfour-fifthsof the actual increase of 12% overthisperiod. The studyfounditplausiblethatthe
upwarddriftinthe proportionof ‘good’honoursclassificationsmaybe due tochangesin
assessmentmethodsandmodularisationof degreeprogrammes.However,Bachanacknowledged
that time-seriesdatausedinthe analysiswas relativelyshortand itwasnot possible tocontrol for
several importantfactors inthe empirical analysisincludingdetaileddescriptionsof learning,
teaching,andassessmentstrategiesemployedatthe course level overalongertime period.One
interestingfindingthatreaffirmspreviousstudieswasthe positive relationshipbetweenpre-entry
UCAS pointsandthe percentage of gooddegreesawarded;a10% increase inmedianUCASentry
pointsincreasesthe share of ‘good’degreesby2.5% ceteris paribus.
In contrast,a studyby JohnesandSoo in2013 found‘negative’ grade inflationrelativetothe year
2005 [17]. The studyfoundthat fewof the time dummieswere significant,albeitwitharisingtrend
between2008 and 2012, where theybecome significantlydifferentfromzero; ceterisparibus,this
correspondedtoanincrease inthe share of ‘good’degreesby0.5% by 2012. The authors note,
withoutimplyingcausality,thatthisincrease coincidedwithaperiodof intensifyingcompetitionand
risingtuitionfeesfrom£3000 peryearin 2006, to £9000 peryear in2012. Interestingly,the authors
testedthe hypothesisthatentrygradesare an importantas a determinantof degree outcomesby
observingthe interactionbetweenentrygradesandthe time-dummiesusedinthe analysis;they
noteda significantpositiveinteractionbetweenbothvariables from2008 onwards,but by2012 this
was statisticallyinsignificantfromzero,indicatinglimitedevidence of grade inflation.However,this
relationshipprovedtobe strongerwhenconsideringasmallersampleof ancient‘elite’universities
comparedto neweruniversities.
‘The jury’s(still) out’maybe aptin describingthe currentacademicpositionongrade inflationatUK
universities.Howemployerswill respondtothe currentmediafrenzyonthistopicisarguablya
more importantissue.Studentsultimatelyenrol inhighereducationtoattainuseful skillswith(inmy
case anyway) the endgoal of improvingtheirfuture earningsprospects,andultimately,theirlife
satisfaction.Iwouldsurmise thatlarge employersare alreadydevisingnew,more in-depthmeansby
whichto evaluate the skills setof future graduates,broadeningtheirrequirementsbeyonddegree
classificationandinterviewperformance alone.Atthe same time,universitiesmustcontinue to
enforce rigorousacademicstandardswhilstmaintaininganeffectivelineof communicationwith
graduate employers.Ileave the readerwith amemorable quote fromthe 2004 superhero film“The
Incredibles”whichIenjoyedwatchinginprimaryschool –in the contextof maintainingacademic
standards,I believe itreallyhitsthe nail onthe head!
“If everyoneis super,no onewill be” – Syndrome,TheIncredibles (2004)
18. Closing Remarks
Overall Ibelievethe three trends presented inthisarticle deservegreaterattentiongoingforward.
There iscertainlya lackof academicresearchfocusedonexplainingwhythese trends have
continuedtostrengthenoverthe past decade.Besidessporadicpressattentiongiventopotential
grade inflationinUKuniversities,there have beenfew opendiscussionsonthese topics.Ashigher
educationisbecomingincreasinglymore accessible,itbecomesevenmore importanttodiscusshow
universitieswillevolve goingforward.Future generationsof prospective studentswill rely inparton
the advice we are obligedtoprovide them, anditisessential thatthisadvice isgroundedfirmlyin
reality.Thiscanonlybe achievedbydiscussing thesetopicsopenly.
References
[1] BIS, Successas a knowledgeeconomy:teaching excellence,socialmobility and studentchoice,
no.May.2016.
[2] Treasury,“Productivityin the UK:The Evidence and the Government’sApproach,”2001.
[3] I. Walker and Y. Zhu,“The Returnsto Education:Evidencefromthe LabourForceSurveys,”
2001.
[4] G. Conlon and P.Patrignani,“TheReturnsto Higher Education Qualifications,” Dep.Bus.
Innov.Ski.,no.45, p. 116, 2011.
[5] UKCES,“Reviewing the RequirementforHigh Level STEM Skills,” 2014.
[6] HESA,“Higher Education StudentEnrolmentsand QualificationsObtained atHigherEducation
Providersin the United Kingdom,”2016.
[7] GovernmentEqualitiesOffice,“Government responseto the Women’sBusinessCouncil,”
2013.
[8] WISE, “Women in Science, Technology,Engineering and Mathematics:TheTalentPipeline
fromClassroomto Boardroom,UKStatistics2014,” vol.W, no.July, p.24, 2015.
[9] S. Coughlan,“UniversityFirst-classDegreesSoaring,”Education &Family,2017.
[10] The CompleteUniversityGuide,“The CompleteUniversity Guide,University LeagueTables
2018,” 2017. [Online].Available:https://www.thecompleteuniversityguide.co.uk/league-
tables/rankings.
[11] The Russell Group,“Funding HigherEducation Part1: Staying on top:The challengeof
sustaining world-classhighereducation in theUK,” Russell Gr. Pap.,no.2,pp. 1–52, 2010.
[12] S. Court,“An analysisof student:staff ratiosand academics’useof time,and potentiallinks
with studentsatisfaction,”Univ.Coll.Union,no.October2012, pp.1–28, 2012.
[13] T. C. U.Guide, “The CompleteUniversity Guide, LeagueTablesMethodology,”2017. [Online].
Available:https://www.thecompleteuniversityguide.co.uk/league-tables/methodology/.
[14] The Independent,“UniversitiesShould ResistGradeInflation WarnsWillets,” 2012.
19. [15] A. Kiureghian and O.Ditlevsen,“Aleatoric or Epistemic? Doesit matter?,” StructSaf.,vol.31,
no.2, pp.105–112, 2009.
[16] R. Bachan, “Gradeinflation in UK highereducation,” Stud.High.Educ.,vol.5079, no.March,
pp.1–21, 2015.
[17] G. Johnesand K.T. Soo,“Grades AcrossUniversitiesOverTime,” 2013.