An assessment of the English and maths skills levels of prisoners in England

Although the direct links between education and reducing recidivism in prisoners are problematic, there is little argument that education is a factor in promoting reintegration and rehabilitation. There is a current focus in prison education on education for employment, and yet there are no recent or unambiguous data about the skills levels of the prison population. The most often quoted figures are both 15 years out of date and deeply flawed in terms of their comparisons with the general population. This article sets out a new study that takes the mandatory initial assessments carried out on every new prisoner between August 2014 and July 2015 and compares them with the national Skills for Life survey conducted in 2011. This provides us with some hard facts about the English and maths skills of the past year's intake of prisoners. The conclusions argue that while the numeracy skills of prisoners are better than previously understood, the cohort has extremely poor literacy skills, and addressing these needs should be a priority for government.

The SFA has contracts with four organizations who provide learning and skills training for prisoners across ten areas of England. Novus (the prisons education department ofThe ManchesterCollege)runsprisoneducationinLondon,theNorthEast,theNorthWest,Kent and Sussex, andYorkshire and Humber; Milton Keynes College runs prison education in the EastMidlands,SouthCentralandtheWestMidlands;WestonCollegeintheSouthWest;and PeoplePlus(formerlyA4E)intheEastofEngland.

Aims of study
The specific aim of the study was to improve our understanding of prisoners' literacy and numeracy levels on entry to prison, and to compare this with the profile of the country as a whole. It has always been clear that this initial study should be seen as the start of that processandprovidesonlythefirstanalysisoftheMAdata,justthosecollectedbytheeducation providersbetweenAugust2014andJuly2015.

How research was conducted
The analysis has been conducted using the provisional data provided by each of the OLASS providers on a prison-by-prison basis.The data only cover English public sector prisons.The officialMAdata,whichareagglomeratedratherthanbrokendownintoindividualprisons,have nowbeenreleasedbyBIS(seebelow),andareinlinewiththeprovisionalfiguresusedhere.

Quantitative data from MAs
The individualized data have been received from 104 prisons. 5 The data come from all four OLASS4providers,andcoveralltypesofprison.Intotaltherearejustover123,000assessment resultsforEnglishandmaths.
Starting with literacy (Table 2), we see that there is a higher percentage of prisoners at everylevelbelowL2.ThedifferencebetweenL2skillsinsideandoutsideprisonisverystark,a differenceofalmost43percentagepoints. Inthegeneralpopulation86percenthaveliteracyskillsatL1orabove,whereasinprison the figure is only 50 per cent. If L1 literacy is considered the appropriate skill level for succeeding in most types of employment,'functional literacy' as it was termed in the Skills for Life programme, then this represents a significant barrier for prisoners looking to gain employmentonrelease. Thenumeracydata(Table3)provideadifferentpattern.Thelowerskillslevelsfornumeracy arequitesimilartothegeneralpopulationandwhile12percentmoreprisonershaveEL3skills, 5percentmorehaveL1skills.TheshortfallatL2isneithersurprisingnoraslargeasforliteracy.
Although it is currently not possible to match MA results with individual prisoners' categorizations,thedatanonethelessillustratethatdifferenttypesofprisonsdohaveprisoners withdifferentprofilesofbasicskills.

MA data by provider
While we received data from all four OLASS 4 providers, the numbers involved were very different(seeTable4). Nonetheless, it is interesting to see if there are any differences in the skills profiles of the fourproviders,andFigure9showsthiscomparison.Despitethedisparitybetweennumbersof assessments,wemightexpectthattheprofileswouldbefairlysimilartoeachother. Somewhat surprisingly, there are quite large differences between providers. For literacy, Novus has 10 percentage points more L1 assessments than PeoplePlus andWeston, while Westonhas5percentagepointsmoreL2assessmentsthanNovus.EL1assessmentsfromMilton KeynesCollegearealmostdoublethoseforManchester(Novus).
There is some disparity when we look at the numeracy assessments, with PeoplePlus's L1beingmarkedlylowerthanthoseofotherproviders,whileitsEL3isthatmuchhigher(see Figure10).

MA data on LDD
As noted above, prisoners self-declare their LDD status. Exactly how this is done, and what adviceandsupportisgiven,willinevitablybedifferentineachprison,butwemaybeableto detectanybiasinthewayprovidersdothisbycomparingthepercentagesofLDDdeclaration byprovider. Figure 11showsadifference ofalmost13percentage points between LDD assessments madebyWestonCollegeandPeoplePlus.
While the overall rate of LDD for all prisons is 32 per cent, the rate of LDD in female prisonsis50percent,asignificantlyhigherfigurethanforthemaleestate.
AlthoughtheprovidershaveprovidedsomefiguresfortypeofLDDdeclared,thesearenot easilycompared.However,itisclearthatdyslexiarepresentsaround60percentofdeclared LDD, with 20 per cent consisting of moderate learning difficulties, the next largest category. Unsurprisingly, evidence suggests that mental health issues are also a significant factor in the prisonpopulation.
Firstly,itisclearthatthehighpercentagesofprisonerswithEntry-levelEnglishskillsare amajor barrier for thosewishing toenter employment on release. Policymakers needto be awarethatadultswiththeselowlevelsofliteracyskillarebyfarthehardesttoaddress,and that sustained educational effort is required for them to achieve functional levels of literacy and numeracy. Such learners would typically require a very thorough and personalized oneto-oneassessmentdesignedtoascertaintowhatextenttheweaknessesarebasedaroundan individual'sliteracyorlanguageabilities,writingorreadingskills,andanyotherrelatedlearning difficultiesormentalhealthproblems.Agreatdealofworkisrequiredtobuildupalearner's self-confidenceandself-esteem,andresultantlearningplansarelikelytobehighlyindividualized. Prisoners with these skills cannot simply be assigned to a regular programme of study in a classroomandexpectedtoprogresswithoutadditionalsupport.
Perhaps less surprisingly, the profile of YOI inmates is closer to that of the general population,withtheexceptionofL2andaboveskills.Thismightsuggestthatonlythelateryears of education have been a problem for this cohort and that they retain most of their school learninginthesebasicskills.
Asexplained,unlessanduntilwecantietheMAdatawithindividualprisoners'records,we cannotprovideanaccurateaccountofhowbasicskillslevelsvaryacrossthedifferentcategories of prisoner or sentence duration.What we have is an indication that prisoners assigned to CategoryA and Category B prisons are particularly weak in their basic skills and in need of specialisthelp.
Although there is a uniformly high level of LDD reported by all categories of prisoner, around32percent,thisisfarhigherforthefemaleestate.Onceagain,thereasonsforthisare notapparentatthisstageandfurtherinvestigationisrequired.
As explained above, the OLASS contracts give providers considerable freedom in how theydelivertheMAs.Whentheresultsacrossproviderswerecompareditwasnotexpected that there would be any great differences between them.This was not the case.There were clear differences in the percentages of prisoners at each level in both literacy and numeracy, the number of cases of LDD recorded, and the mean number of assessments recorded per prison.Thismaywellbecausedbytheprovidershavingdifferentsizesandcategoriesofprison, butcouldalsobecausedbyproviderstakingdifferentapproachestothecollectionofdata.If practicebecomesmorediverseinthefuture,forinstancethroughusingdifferentIAs,thenthese differenceswillincreaseandcomparisonsbecomeevenmoreproblematic.
Another area where we have been unable to produce any conclusions is for prisoners forwhomEnglishisnottheirfirstlanguage(ESOLlearners).Thisinformationisnotcurrently collectedintheMAdata,butsomeindicatorsmaybegainedwhentheILRdataarereleasedby cross-referencingwithESOLqualificationsenrolments.
Thisdatacollectionandanalysisneedstoberepeatedoveranumberofconsecutiveyears inordertobuildupasolidsetofdataonprisoners'literacyandnumeracylevels.Furtherwork needs to be carried out to understand better the language and literacy implications of the skillsprofiles,theimpactofdifferentproceduresbetweenprovidersandbetweenprisons,and howprisonersself-declaretheirLDDstatus.Ideallywewouldliketofindwaysoflinkingthese educationdatawiththeprisondataaboutprisonsentencesandcategory.