Are Secondary Schools Imparting Digital Skills? An Empirical Assessment

Abstract

The aim of this article was to analyse the employability of currently enrolled secondary education (classes 9–12) students aged 14–21, specifically their functional digital skills. Digital skills are increasingly being recognised as a key foundational skill that also enhances employability. Gaining digital skills at the secondary education is important because it is one of the foundational skills that help prepare students transition to work and prepares them for life. There were three major objectives in this article: (a) to examine the trends and transitions in the acquisition of functional digital skills of currently enrolled secondary school Indian students according to their socio-economic and demographic profiles; (b) empirically investigate the factors influencing the acquisition of functional digital skills in urban areas and (c) examine whether the policy of providing schools with computers has had any discernible impact on the acquisition of functional digital skills of these students. Using the National Statistical Office 2017–2018 data on expenditure on education, we found that at the national level only 42% of the enrolled secondary school students had the ability to operate a computer and 46% had the ability to browse Internet in 2017–2018. The attainment of functional digital skills differed across rural and urban regions. The individual characteristics, socio-economic profile of households and school-related indicators were factors that explained the likelihood of students’ being equipped with functional digital skills. A key result was that students who had digital devices at home were more likely to have functional digital skills. The government had introduced computers in secondary schools in 2004 in India. We found evidence of a positive association between the provision of functional computers at secondary schools and attainment of digital skills, even for students from households with no computers at home. A key policy recommendation is that providing computers at schools can help overcome barriers to access to digital devices at home and improve the attainment of digital skills.

JEL Codes: I2, O15, O3

Keywords

Secondary Education Indian Schooling Digital Skills ICT Skills India Employability

1. Introduction

The key challenge in India is one of employability, that is, to ensure educated people have the appropriate set of skills to make a successful transition to a life of work (NCAER, 2018). Secondary school students (classes 9–12) need to have foundational skills that enhance employability as this prepares them for life when they complete school. Information and Communication Technology (ICT) skills, that is, skills which can range from a basic ability to use digital devices and tools in daily life to programming, coding and computational thinking are a key part of that foundational skills set (GoI, 2020; NCAER, 2018). UNESCO and Intel (2017) have also said that digital skills are important for both life and work, given that digital devices are omnipresent.

There were three major objectives in this article: (a) to examine the trends and transitions in the acquisition of functional digital skills of currently enrolled secondary school Indian students according to their socio-economic and demographic profiles; (b) empirically investigate the factors influencing the acquisition of functional digital skills in urban areas and (c) to examine whether the policy of providing schools with computers has had any discernible impact on the acquisition of functional digital skills of these students.

Digital skills have not been examined in the Indian empirical literature partly because data was not available till 2014. Ocansey and Sharma (2020) partially dealt with this in their recent article on the digital readiness of schools in Maharashtra in terms of their digital infrastructure, support, research and skills and so on. Since ours is one of the first articles in this area, defining the terminology becomes important—ICT skills and digital skills have different meanings. ICT skills involve the use of digital devices and online applications, whereas digital skills indicate a broader concept of ICT skills plus cognitive and socio-emotional skills. ICT skills may be defined as functional digital skills and are being increasingly recognised as a foundational skill along with reading, writing and numeracy (UNESCO, 2018).¹ Evidence indicates that ICT skills improve employability and can pay higher returns on the labour market (Atasoy, 2011; Falck et al., 2016; Garrido et al., 2012; Rani et al., 2019). A key point made by UNESCO and Intel (2017) and Garrido et al. (2012) was that access to ICT devices is important for acquiring and using digital skills.

Secondary education is the appropriate place to assess acquisition of functional ICT skills as it ‘completes the provision of basic education that began at the primary level and aims at laying the foundations for lifelong learning and human development, by offering more subject- or skill-oriented instruction using more specialised teachers’ (World Development Indicators website). This is important for India for three reasons: (a) The relatively low rate of transition between the lower secondary classes (9 and 10) and upper secondary classes (11 and 12); (b) the fact that youth do not necessarily continue to study further after completing their secondary education and (c) high youth unemployment rates after finishing school.² Further, the ASER (2018) found that 17.6% of young people (aged 14–18) had never used a mobile phone, 63.7% had never used the Internet and 59.7% had never used a computer.

Using data from the National Statistical Office (NSO) (the erstwhile National Sample Survey Office, NSSO) for 2014 (GoI, 2016) and 2017–2018 (GoI, 2019d), we found an improvement in access and attainment of functional digital skills of currently enrolled secondary school students between 2014 and 2017–2018. At the national level in 2017–2018, only 42% of the enrolled secondary students had the ability to use a computer and 46% had the ability to browse the Internet. The attainment of functional digital skills differed across rural and urban regions. A key result was that students who had digital devices at home were more likely to have functional digital skills. Using a logistic model for urban students, we found evidence of a positive association between the provision of functional computers at secondary schools and attainment of digital skills, even for those students belonging to households with no access to computers at home. Providing schools with computers can overcome barriers to access of digital devices at home and thereby improve the attainment of digital skills. This is an important result because Ocansey and Sharma (2020) found that schools in Maharashtra relied on corporate social responsibility for the provision of devices in schools. Access to Internet in schools depended on teachers’ willingness to use their own mobile data. The remainder of the article is organised as follows. Section 2 gives a theoretical motivation for this work. Section 3 describes the data. Section 4 presents the data description and Section 5 describes the empirical methodology. Section 6 presents the major findings. The last section presents the conclusions and discusses some of the major policy implications.

2. Theoretical Motivation

In order to be employable, functional digital skills should be testable, certifiable and recognisable by an employer. Therefore, institutional access to ICT skills is important, and this is provided either through schools or private institutions, though the latter may involve a cost.³ The ownership of computers at home gives students an added advantage.⁴ However, the relatively low ownership of computers in India means that access has to be provided through alternative means.⁵ Therefore, the most scalable and equitable access for students to acquire certifiable ICT skills remains at school. The key empirical hypothesis in this article is whether publicly provided computers in schools have helped students acquire functional digital skills.

3. Data

This article utilised unit-level information from household surveys (known as the National Sample Survey [NSS]) conducted by the NSO.⁶ NSS surveys are nationally representative household surveys and follow stratified multi-stage sampling design. It conducted a survey on Household Social Consumption: Education in July 2017–June 2018 (the 75th Round). Earlier surveys on the same subject were carried out in 2007–2008 (the 64th Round, GoI, 2010) and 2014 (the 71st Round, GoI, 2016), but ICT indicators have been collected only in 2014 and 2017–2018.

Figure 1

Source: Authors’ computations from GoI (2016).

Figure 2

Source: Authors’ computations from GoI (2019d).

The empirical analysis included only those students aged 14–21 who were currently enrolled in secondary and senior secondary schools and diploma/certificate courses equivalent to the secondary and senior secondary levels of education.⁷ This group accounted for 91% and 92% of secondary students in 2014 and 2017–2018, respectively (Figures 1 and 2).

Functional digital skills are measured in terms of ability to operate computer and ability to browse Internet. While both the NSS rounds were comparable from the statistical point of view, the measurement of access to digital devices and functional digital skills differed substantially between the two rounds (GoI, 2019d) as detailed in Table 1.

4. Data Description

Of the currently enrolled students in secondary and senior secondary school aged 14–21, in Round 75, 70.7% of this sample was from the rural areas and 57.5% were males (the corresponding figures in Round 71 were 70.2% and 56.3%, respectively).⁸ The two rounds are not directly comparable (see Table 1). The key messages from Table 2 were: those in urban areas had better access to digital infrastructure; despite a low household ownership of computers, Internet access was higher than computer ownership in both periods; and students had better access to ICT devices if they belonged to the higher expenditure quintiles. NSO surveys provide monthly per capita consumption expenditure (MPCE) of surveyed households and all households are divided into various classes with equal representation of each MPCE Class.

Tables 3 and 4 present the functional digital skills of secondary students based on their access to digital devices at home. The tables illustrate the substantial differences in students’ functional digital skills depending on home ownership/access to digital devices. They also highlight the differences in gender attainment of functional digital skills.

A higher percentage of students living in urban areas had attained functional digital skills compared to students living in rural areas (Table 5). Further, the share of students who could browse the Internet showed a marked improvement from 34.2% in 2014 to 46.8% in 2017–2018. The share of students who had actually used Internet in the last 30 days was lower than the share that had Internet browsing ability in 2017–2018 (see Table 1 for definition). More than twice the share of urban students had used the Internet (60.5%) than rural students (28.8%) in 2017–2018.

Table 1
Differences between NSS Rounds 71 and 75

S. No. Item Round 71 2014 Round 75 2017–2018

1. Collection of information on current enrolment/attendance, education particulars of basic course and expenditure on education, 5–29 ages 3–35 age

2. Age for ICT skills Age 14 and above Age 5 and above

3. Ownership of computer—the term ‘computer’ Computer includes a desktop, laptop, palmtop, notebook, netbook, smartphone, tablets, etc. Computer here includes devices like a desktop, laptop, notebook, netbook, palmtop, tablet (or similar handheld devices). It did not include a smartphone.

4. Access to use Internet facility Whether any member of the household (aged 14 and above) has access to an Internet facility Whether the household has an Internet facility

5. Able to use computer For word processing/typing Able to operate a computer for copying or moving a file or folder, using copy and paste tools to duplicate or move information within a document, sending e-mails with attached files, using basic arithmetic formulae in a spreadsheet, connecting and installing new devices, finding, downloading, installing and configuring software, creating electronic presentations with presentation software, transferring files between a computer and other devices, writing a computer program using a specialised programming language.

6. Able to use Internet Able to use Internet for searching for desired information. Whether able to use Internet: household member was able to use Internet browser for website navigation, using e-mail and social networking applications, etc., to find, evaluate and communicate information.

7. Actually used Internet Able to use Internet for sending e-mails Whether actually used Internet during last 30 days

S. No.	Item	Round 71 2014	Round 75 2017–2018
1.	Collection of information on current enrolment/attendance, education particulars of basic course and expenditure on education,	5–29 ages	3–35 age
2.	Age for ICT skills	Age 14 and above	Age 5 and above
3.	Ownership of computer—the term ‘computer’	Computer includes a desktop, laptop, palmtop, notebook, netbook, smartphone, tablets, etc.	Computer here includes devices like a desktop, laptop, notebook, netbook, palmtop, tablet (or similar handheld devices). It did not include a smartphone.
4.	Access to use Internet facility	Whether any member of the household (aged 14 and above) has access to an Internet facility	Whether the household has an Internet facility
5.	Able to use computer	For word processing/typing	Able to operate a computer for copying or moving a file or folder, using copy and paste tools to duplicate or move information within a document, sending e-mails with attached files, using basic arithmetic formulae in a spreadsheet, connecting and installing new devices, finding, downloading, installing and configuring software, creating electronic presentations with presentation software, transferring files between a computer and other devices, writing a computer program using a specialised programming language.
6.	Able to use Internet	Able to use Internet for searching for desired information.	Whether able to use Internet: household member was able to use Internet browser for website navigation, using e-mail and social networking applications, etc., to find, evaluate and communicate information.
7.	Actually used Internet	Able to use Internet for sending e-mails	Whether actually used Internet during last 30 days

Source: GoI (2016b) and GoI (2019d).

Table 2

Currently Enrolled Students in Secondary School with Access to ICT Devices (%)

	NSS Round 71 (2014)			NSS Round 75 (2017–2018)
	Computer	Internet	Both Computer and Internet	Computer	Internet	Both Computer and Internet
Overall	19	39	17	13	31	10
Rural	11	28	9	7	23	5
Urban	38	64	36	27	49	23
Expenditure Quintile-wise
Q1	4	16	4	6	17	4
Q2	9	25	7	6	22	5
Q3	15	35	13	10	28	8
Q4	21	43	19	14	35	12
Q5	37	62	35	26	48	22

Source: Authors’ computations from GoI (2016; 2019d).

Note: Monthly household expenditure was used to divide the entire population into five equal groups. The bottom-most quintile (20%) is labelled as Q1 and so on.

Table 3

Currently Enrolled Secondary School Students with ICT Skills, Based on ICT Access, 2014 (%)

	Able to Operate Computer for Word Processing/Typing			Able to Use Internet			Able to Send E-Mails			ICT Skills
	M	F	T	M	F	T	M	F	T	M	F	T
All students	42	33	38	38	29	34	36	24	31	31	20	26
Students with ICT Skills in Households with a Computer
Yes	88	79	84	81	74	78	82	69	76	73	60	67
No	31	23	27	28	20	24	25	14	20	20	12	16
Students with ICT Skills in Households with Both Computer and Access to Internet
Yes	90	81	86	84	76	80	86	73	80	76	63	71
No	32	23	28	28	20	25	25	15	20	21	12	17

Source: Authors’ computations from GoI (2016).

Note: T denotes total, M denotes males and F denotes females.

Table 4

Currently Enrolled Secondary School Students with ICT Skills, Based on ICT Access, 2017–2018 (%)

	Able to Operate Computer			Able to Use Internet			Used Internet in Last 30 Days
	M	F	T	M	F	T	M	F	T
All students	46.8	37.0	42.7	52.7	38.8	46.8	44.0	30.1	38.1
Students with ICT Skills in Households with a Computer
Yes	91.8	85.5	89.3	90.9	82.2	87.5	85.3	75.1	81.4
No	39.9	30.7	35.9	46.8	33.2	40.9	37.6	24.2	31.8
Students with ICT Skills in Households with Both Computer and Internet
Yes	92.8	85.9	90.2	93.5	84.7	90.2	90.1	80.7	86.5
No	41.0	31.9	37.1	47.6	34.0	41.7	38.2	24.8	32.4

Source: Authors’ computations from GoI (2019d).

Note: T denotes total, M denotes males and F denotes females; Unlike the 2014 Survey, the three indicators of ICT skills cannot be combined into a single measure.

Table 5

Currently Enrolled Secondary School Students with ICT Skills, Rural–Urban (%)

Sector	NSS Round 71 (2014)				NSS Round 75 (2017–2018)
Sector	Can Use Computer for Word Processing/Typing	Internet Browsing Ability	Mail Sending Ability	ICT Skills	Can Use Computer	Internet Browsing Ability	Used Internet during Last 30 Days
Rural	27.3	23.7	20.5	16.9	32.4	37.3	28.8
Urban	63.8	58.9	55.1	47.7	67.3	69.7	60.5
Total	38.1	34.2	30.8	26.1	42.7	46.8	38.1

Source: Authors’ computations (from GoI, 2016b, 2019d).

Female students lag behind their male counterparts in the attainment of all types of functional digital skills, irrespective of whether digital devices were available at home or not. Other findings were that the ownership of digital devices has a positive link with attainment of functional digital skills, and that there were significant rural–urban differences in the attainment of functional digital skills among students.

5. Empirical Methodology

Only data from Round No. 75 are used to analyse the factors that explain the likelihood of students attaining functional digital skills as definitions are not consistent across both the rounds (see Table 1). Second, given the rural–urban difference in access to digital devices and functional digital skill attainment gaps,⁹ we focussed our empirical analysis on urban areas. Ocansey and Sharma (2020) point out that urban schools in Maharashtra were better equipped with infrastructure and teachers for imparting digital skills. Focusing the analysis on just students from urban areas also rules out endogeneity between schools and attainment of skills of a particular region to some extent. Third, our empirical analysis covered all the states and Union Territories (UTs)s in India, as in the survey, but for ease of comparison, the Northeastern states and UTs had been clubbed into two separate categories.¹⁰

The first step was to analyse the factors which explain the likelihood of students acquiring functional digital skills. Three measures of functional digital skills were used as dependent variables—the ability to operate a computer, ability to browse the Internet, and ability to operate a computer and browse Internet. The explanatory variables were clubbed into four: (a) individual characteristics; (b) household socio-economic characteristics; (c) school characteristics and (d) state dummy variables.

Three variables were included in the individual characteristics—age, gender and level of study (senior secondary). As males showed higher attainment of functional digital skills (Table 5), a positive sign was expected on this explanatory variable. The impact of age on functional digital skills was ambiguous as younger children may be more digitally savvy. There has been a rise in the use of digital technology in younger children (Burns & Gottschalk, 2019, Ch. 2). However, older children may have had more years of formal digital training in school or gained experience from working. Given that a large number of children in secondary school were outside the standard age group of 14–18, this was an important variable to consider. The same intuition would apply to the third variable, the level of study, that is, students enrolled in senior secondary grades 11 and 12.

The household characteristics comprised five variables—size, occupation, social group, expenditure quintile group and ownership of/access—to digital devices at home. An increase in household size would have a negative impact on the acquisition of digital skills. Households with smaller families would be able to invest more in their children (Becker & Lewis, 1973), a finding that has been confirmed by Indian empirical evidence (Kugler & Kumar, 2017).

There were four household occupation types in urban areas—self-employed, regular wage/salary, casual labour and others.¹¹ Intuitively, relative to self-employed, casual labour and other households, the regular wage/salaried household would offer a more stable and better home environment in terms of access to facilities, support, encouragement (Mudassir & Abubakar, 2015), and social capital (Myroniuk et al., 2017). Therefore, one would expect a positive sign on this household type, that is, secondary school students from regular wage/salary households would be more likely to have functional digital skills.

There were four social household groups—General, Scheduled Castes (SCs), Scheduled Tribes (STs) and Other Backward Castes. Mukherjee et al. (2016) documented the relatively lower digital access of SCs and STs. NCERT (2015) and Young Lives (2017) also have documented the relatively lower learning outcomes of children from SC/ST households. Therefore, the probability of a secondary school student having functional digital skills would be higher if they were from a general background compared to others, that is, one would expect a positive sign on this variable. The likelihood of a secondary school student possessing functional digital skills would be higher if the student belonged to a higher expenditure quintile group. The expenditure variable acted as a proxy for income. The likelihood that a student possesses digital skills would be higher in a household which owned digital devices (Table 4). Households without a computer or Internet were given the value of one; therefore, we would expect a negative sign.

Figure 3

Source: Authors’ computations from GoI (2019d).

School characteristics include two variables—school type (government/private) and whether English was a medium of instruction. It was expected that students enrolled in private schools were more likely to have digital skills as these schools tended to have better infrastructure (Thakur, 2014). Since government schools had been classified as one, one would expect a negative sign on this variable. The impact of having English as a medium of instruction on the probability of secondary school students acquiring functional digital skills was positive. Since English was largely the terminology used to impart digital skills, better knowledge of the language may improve the chances of acquiring functional digital skills.

The state dummy variables were used to capture any specific state-effects or state policies. Education being a part of the Concurrent list, state governments have the primary role in managing school-level education in terms of curriculum, teachers, textbooks, etc. The state dummy variable would presumably capture all these variations. A state-wise distribution of student with digital skills (Figure 3) showed that100% of currently enrolled secondary school students in Goa in the age group 14–21 had functional digital skills.

6. Empirical Results

6.1 Factors Affecting Digital Skills

The first empirical research question is what factors explain the acquisition of digital skills by urban secondary school students. A logistic regression was performed to ascertain the impact of individual, household and school-related characteristics on the likelihood of secondary school students having digital skills. We ran three separate logistic regression models to test three different dependent variables, each converted into binary codes:

Model 1: Ability to operate computer/computer skills (1 if yes, 0 otherwise).

Model 2: Internet browsing ability/Internet skills (1 if yes, 0 otherwise).

Model 3: Ability to operate computer and browse the Internet/ICT skills (1 if yes, 0 otherwise).

The empirical results were reported in Table 6.¹² All three regressions were statistically significant, as the P values associated with the chi square test was <0.05. The models correctly identified 78.9%, 80.1% and 77.6% of students with the ability to operate a computer, ability to browse the Internet and ability to do both, respectively. The pseudo-R² explained 44.1%, 44.7% and 44.5% of the variance in the three regressions—ability to operate a computer, ability to browse the Internet and ability to do both, respectively.¹³ We first checked to see if the signs for the three regressions matched with our reasoning, but interpret the results only for the third model, where a student was likely to have both computer and Internet skills.

Table 6
Factors Affecting Ability of Secondary School: Students to Operate a Computer

I- Ability to Operate Computer
II-Ability to Browse Internet
III-Ability to Operate Computer and Browse Internet

B Odds Ratio B Odds Ratio B Odds Ratio

Individual

Gender (Male = 1) 0.29*
[0.001] 1.34 0.39*
[0.001] 1.47 0.37*
[0.001] 1.45

Age 0.14*
[0.001] 1.15 0.20*
[0.001] 1.22 0.17*
[0.001] 1.19

Level of study (Sr. Secondary) 0.62*
[0.002] 1.86 0.77*
[0.002] 2.16 0.68*
[0.001] 1.98

Household Characteristics

Household size −0.12*
[0.000] 0.88 −0.13*
[0.000] 0.88 −0.11***
[0.000] 0.90*

Household Occupation (Control = Self-employed)

Regular wage/salary 0.05*
[0.001] 1.05 0.01*
[0.002] 1.01 0.051*
[0.001] 1.05

Casual labour −0.41*
[0.002] 0.67 −0.35*
[0.002] 0.70 −0.33*
[0.002] 0.72

Others −0.03*
[0.002] 0.97 0.14*
[0.003] 1.15 0.11*
[0.002] 1.11

Social Group (Control = General)

ST −0.77*
[0.003] 0.46 −0.56*
[0.003] 0.57 −0.70*
[0.003] 0.50

SC −0.31*
[0.002] 0.74 −0.40*
[0.002] 0.67 −0.27*
[0.002] 0.76

OBC −0.36*
[0.002] 0.70 −0.34*
[0.002] 0.71 −0.38*
[0.001] 0.68

Urban Household Expenditure Quintile (Control = Highest Q5)

Bottom Q1 −0.76*
[0.003] 0.47 −0.66*
[0.003] 0.52 −0.85*
[0.002] 0.43

Q2 −0.61*
[0.002] 0.54 −0.46*
[0.002] 0.63 −0.61*
[0.002] 0.54

Q3 −0.34*
[0.002] 0.71 −0.38*
[0.002] 0.68 −0.41*
[0.002] 0.66

Q4 −0.44*
[0.002] 0.64 −0.40*
[0.002] 0.67 −0.47*
[0.002] 0.62

Household without computer (=1) −1.92*
[0.002] 0.15 −0.99*
[0.002] 0.37 −1.58*
[0.002] 0.21

Household without Internet (=1) −0.87*
[0.001] 0.42 −1.71*
[0.002] 0.18 −1.06*
[0.001] 0.35

School Characteristics

School type (Government) −0.15*
[0.002] 0.86 −0.16*
[0.002] 0.85 −0.13*
[0.001] 0.88

Medium of instruction (English = 1) 0.87*
[0.002] 2.38 0.76*
[0.002] 2.13 0.84*
[0.002] 2.30

States (Control = Northeastern States)

Andhra Pradesh −0.07*
[0.006] 0.93 -0.05*
[0.006] 0.95 –0.12*
[0.006] 0.89

Assam 0.94*
[0.008] 2.56 0.93*
[0.009] 2.54 0.86*
[0.008] 2.36

Bihar 0.81*
[0.006] 2.25 0.881*
[0.007] 2.41 0.88*
[0.006] 2.40

Chhattisgarh 0.16*
[0.007] 1.17 −0.19*
[0.007] 0.82 0.083*
[0.007] 1.09

Goa 4.91*
[0.083] 135.38 4.73*
[0.091] 112.71 5.14*
[0.083] 170.79

Gujarat 1.71*
[0.006] 5.5 0.88*
[0.006] 2.42 1.30
[0.006] 3.66

Haryana 0.53*
[0.006] 1.70 0.82*
[0.007] 2.27 0.51*
[0.006] 1.67

Himachal Pradesh 1.30*
[0.016] 3.67 0.13*
[0.014] 1.14 0.74*
[0.014] 2.10

Jammu & Kashmir −0.37*
[0.008] 0.69 −0.36*
[0.009] 0.70 −0.28*
[0.008] 0.76

Jharkhand 0.93*
[0.007] 2.53 0.45*
[0.007] 1.56 0.92*
[0.007] 2.5

Karnataka 1.04*
[0.006] 2.84 0.64*
[0.006] 1.89 1.092*
[0.006] 2.98

Kerala 3.4*
[0.007] 29.82 2.0*
[0.007] 7.36 2.42*
[0.006] 11.28

Maharashtra 1.22*
[0.005] 3.38 0.76*
[0.006] 2.13 1.136*
[0.005] 3.11

Madhya Pradesh 0.52*
[0.006] 1.67 0.40*
[0.006] 1.50 0.60*
[0.006] 1.83

Odisha 0.50*
[0.007] 1.65 0.22*
[0.007] 1.24 0.39*
[0.007] 1.47

Punjab 1.71*
[0.007] 5.51 1.85*
[0.008] 6.34 1.66*
[0.007] 5.23*

Rajasthan 0.52*
[0.006] 1.68 −0.001
[0.006] 1.0 0.302*
[0.006] 1.35

Telangana −0.36*
[0.006] 0.70 0.12*
[0.006] 1.12 −0.21*
[0.006] 0.81

Tamil Nadu 1.21*
[0.006] 3.36 0.55*
[0.006] 1.73 0.99***
[0.006] 2.69*

Uttar Pradesh 0.14*
[0.006] 1.15 0.06*
[0.006] 1.06 0.16*
[0.006] 1.170

Union Territories 1.82*
[0.006] 6.15 1.81*
[0.007] 6.13 1.79*
[0.006] 5.97

Uttarakhand 0.78*
[0.009] 2.18 0.87*
[0.01] 2.39 0.92***
[0.009] 2.52*

West Bengal 1.17*
[0.006] 3.22 0.89*
[0.006] 2.44 1.14***
[0.006] 3.12*

Constant 0.39*
[0.010] 1.48 −0.51*
[0.011] 0.60 −0.619***
[0.010] (0.538)

No of observations 17,154 17,154 17,154

Wald Chi² 12838.2 19181.4 24852.8

Overall percentage of cases correctly predicted by the model 78.9 80.1 77.6

Pseudo-R² (Nagelkerke) 0.441 0.447 0.445

	I- Ability to Operate Computer	II-Ability to Browse Internet	III-Ability to Operate Computer and Browse Internet
Individual
Gender (Male = 1)	0.29*** [0.001]	1.34**	0.39*** [0.001]	1.47**	0.37*** [0.001]	1.45**
Age	0.14*** [0.001]	1.15**	0.20*** [0.001]	1.22**	0.17*** [0.001]	1.19**
Level of study (Sr. Secondary)	0.62*** [0.002]	1.86**	0.77*** [0.002]	2.16**	0.68*** [0.001]	1.98**
Household Characteristics
Household size	−0.12*** [0.000]	0.88**	−0.13*** [0.000]	0.88**	−0.11*** [0.000]	0.90*
Household Occupation (Control = Self-employed)
Regular wage/salary	0.05*** [0.001]	1.05	0.01*** [0.002]	1.01	0.051*** [0.001]	1.05
Casual labour	−0.41*** [0.002]	0.67**	−0.35*** [0.002]	0.70	−0.33*** [0.002]	0.72**
Others	−0.03*** [0.002]	0.97	0.14*** [0.003]	1.15	0.11*** [0.002]	1.11**
Social Group (Control = General)
ST	−0.77*** [0.003]	0.46**	−0.56*** [0.003]	0.57**	−0.70*** [0.003]	0.50**
SC	−0.31*** [0.002]	0.74**	−0.40*** [0.002]	0.67**	−0.27*** [0.002]	0.76**
OBC	−0.36*** [0.002]	0.70**	−0.34*** [0.002]	0.71**	−0.38*** [0.001]	0.68**
Urban Household Expenditure Quintile (Control = Highest Q5)
Bottom Q1	−0.76*** [0.003]	0.47**	−0.66*** [0.003]	0.52**	−0.85*** [0.002]	0.43**
Q2	−0.61*** [0.002]	0.54**	−0.46*** [0.002]	0.63**	−0.61*** [0.002]	0.54**
Q3	−0.34*** [0.002]	0.71**	−0.38*** [0.002]	0.68**	−0.41*** [0.002]	0.66**
Q4	−0.44*** [0.002]	0.64**	−0.40*** [0.002]	0.67**	−0.47*** [0.002]	0.62**
Household without computer (=1)	−1.92*** [0.002]	0.15**	−0.99*** [0.002]	0.37**	−1.58*** [0.002]	0.21**
Household without Internet (=1)	−0.87*** [0.001]	0.42**	−1.71*** [0.002]	0.18**	−1.06*** [0.001]	0.35**
School Characteristics
School type (Government)	−0.15*** [0.002]	0.86**	−0.16*** [0.002]	0.85**	−0.13*** [0.001]	0.88**
Medium of instruction (English = 1)	0.87*** [0.002]	2.38**	0.76*** [0.002]	2.13**	0.84*** [0.002]	2.30**
States (Control = Northeastern States)
Andhra Pradesh	−0.07*** [0.006]	0.93**	-0.05*** [0.006]	0.95**	–0.12*** [0.006]	0.89**
Assam	0.94*** [0.008]	2.56**	0.93*** [0.009]	2.54**	0.86*** [0.008]	2.36**
Bihar	0.81*** [0.006]	2.25**	0.881*** [0.007]	2.41**	0.88*** [0.006]	2.40**

Chhattisgarh	0.16*** [0.007]	1.17**	−0.19*** [0.007]	0.82**	0.083*** [0.007]	1.09**
Goa	4.91*** [0.083]	135.38**	4.73*** [0.091]	112.71**	5.14*** [0.083]	170.79**
Gujarat	1.71*** [0.006]	5.5**	0.88*** [0.006]	2.42**	1.30 [0.006]	3.66**
Haryana	0.53*** [0.006]	1.70**	0.82*** [0.007]	2.27**	0.51*** [0.006]	1.67**
Himachal Pradesh	1.30*** [0.016]	3.67**	0.13*** [0.014]	1.14**	0.74*** [0.014]	2.10**
Jammu & Kashmir	−0.37*** [0.008]	0.69**	−0.36*** [0.009]	0.70**	−0.28*** [0.008]	0.76**
Jharkhand	0.93*** [0.007]	2.53**	0.45*** [0.007]	1.56**	0.92*** [0.007]	2.5**
Karnataka	1.04*** [0.006]	2.84**	0.64*** [0.006]	1.89**	1.092*** [0.006]	2.98**
Kerala	3.4*** [0.007]	29.82**	2.0*** [0.007]	7.36**	2.42*** [0.006]	11.28**
Maharashtra	1.22*** [0.005]	3.38**	0.76*** [0.006]	2.13**	1.136*** [0.005]	3.11**
Madhya Pradesh	0.52*** [0.006]	1.67**	0.40*** [0.006]	1.50**	0.60*** [0.006]	1.83**
Odisha	0.50*** [0.007]	1.65**	0.22*** [0.007]	1.24**	0.39*** [0.007]	1.47**
Punjab	1.71*** [0.007]	5.51**	1.85*** [0.008]	6.34**	1.66*** [0.007]	5.23*
Rajasthan	0.52*** [0.006]	1.68**	−0.001 [0.006]	1.0	0.302*** [0.006]	1.35**
Telangana	−0.36*** [0.006]	0.70**	0.12*** [0.006]	1.12**	−0.21*** [0.006]	0.81**
Tamil Nadu	1.21*** [0.006]	3.36**	0.55*** [0.006]	1.73**	0.99*** [0.006]	2.69*
Uttar Pradesh	0.14*** [0.006]	1.15**	0.06*** [0.006]	1.06**	0.16*** [0.006]	1.170**
Union Territories	1.82*** [0.006]	6.15**	1.81*** [0.007]	6.13**	1.79*** [0.006]	5.97**
Uttarakhand	0.78*** [0.009]	2.18**	0.87*** [0.01]	2.39**	0.92*** [0.009]	2.52*
West Bengal	1.17*** [0.006]	3.22**	0.89*** [0.006]	2.44**	1.14*** [0.006]	3.12*
Constant	0.39*** [0.010]	1.48	−0.51*** [0.011]	0.60	−0.619*** [0.010]	(0.538)
No of observations	17,154		17,154		17,154
Wald Chi²	12838.2		19181.4		24852.8
Overall percentage of cases correctly predicted by the model	78.9		80.1		77.6
Pseudo-R² (Nagelkerke)	0.441		0.447		0.445

Source: Authors’ computations from GoI (2019a; 2019d).

Notes: (a) The model is based on urban students aged 14–21 years currently enrolled in Grades 9–12; (b) the model shows weighted results; (c) ***, ** and * indicate 1%, 5% and 10% levels of statistical significance, respectively and (d) the significance of the odds ratio is determined using the 95% confidence interval. If the interval includes one, then it is considered statistically insignificant.

Individual characteristics

Gender: As expected males had higher functional digital skills in all the three dependent variables. The odds of a male student having both the ability to operate a computer and browse the Internet was 1.45 times higher than that of a female, ceteris paribus.

Age: This was positive in all the regressions: older students were 1.2 times more likely to have both the ability to operate a computer and the Internet, ceteris paribus.

Level of study: The sign was positive in all the regressions: the odds of senior secondary student having the ability to operate both a computer and the Internet was 1.98 times higher than secondary level student, ceteris paribus.

Household characteristics

Household size: As expected this had a negative sign. A unit increase in their household size would lower the odds of a student possessing the ability to operate a computer and browse the Internet by 10.4%, ceteris paribus. This fit right into the quantity-quality trade-off (Becker & Lewis, 1973).

Household occupation: While regular salaried households had the intuitively expected positive sign, the coefficients were not statistically significant in any of the three regressions. The sign for casual labour households was negative, as expected; thus, the odds of a student from a casual labour household having computer and Internet skills was 27.9 times lower than if they came from a self-employed household, ceteris paribus. In contrast, a student from a household type classified as ‘others’ was 11.4 times more likely to have computer and Internet skills compared to a student from a self-employed household. In the other two regressions, the coefficient on ‘others’ was not statistically significant. In sum, the odds of a student from a casual wage labour household possessing functional digital skills were lower.

Household social group: The signs for households classified as SCs, STs and OBCs were expected to be negative: the odds of a student from an ST household possessing the ability to operate both a computer and the Internet was 0.5 times or 50% lower than a student from a general background, ceteris paribus. The corresponding numbers for SCs and OBCs were 23% and 31%, respectively.

Household expenditure quintile group: As expected, all the household expenditure quintile groups had negative signs because the control was the highest expenditure quintile group. This was true for all three regressions. The odds of a student from the lowest expenditure quintile group possessing the ability to operate both a computer and the Internet was 0.43 times or 57.2% lower than a student from the highest quintile group, ceteris paribus. The corresponding numbers for the second, third and fourth quintile groups were 45.9%, 33.8% and 37.7%, respectively.

Household with computers and Internet: As expected the sign on both the coefficients were negative. The odds of a student from a household without computer access having both the ability to operate computer and browse Internet was 0.21 times (79.4%) lower than that of a student from a household with access to computers, ceteris paribus. The corresponding number for households with Internet was 65.4%. Clearly access to digital devices was important for any kind of attainment of digital skills.

Additional robustness checks were carried out to test for multicollinearity, dropping one household variable at a time. The results were broadly similar with pseudo-R² reducing from the baseline. Overall, the results remain the same and can be made available on request.

School characteristics

Type of school: As expected, the sign was negative in all the regressions. The odds of a student enrolled in a government school having both the ability to operate a computer and browse the Internet was 0.88 times (12.1%) lower than that of a student from a private school, ceteris paribus.

Medium of education: Within school-related indicators, medium of instruction showed a strong association with the acquisition of digital skills: the sign was positive in all three regressions. The odds of a student enrolled in an English medium school having both the ability to operate a computer and browse the Internet was 2.3 times (113.4%) higher than that of a student from a non-English medium school, ceteris paribus.¹⁴

State effects

Barring the Northeast all the states’ dummy variables were included in each of the three regressions. Therefore, the signs on the dummy variables were relative to the Northeast and had to be interpreted as such. All the signs were consistent across all the three regressions barring a few. Three states had negative signs, namely, Andhra Pradesh, Jammu & Kashmir and Telangana in all the three regressions. The odds of a student living in Goa having both the ability to operate a computer and browse the Internet was 170.8 times higher than that of a student from the Northeast, ceteris paribus, for the third regression (dependent variable was students having the ability to both operate a computer and browse Internet). The odds ratio for Kerala was 11.3 (Table 6). In Rajasthan, the odds ratio was not statistically significant in the second regression, (dependent variable was students having the ability to browse Internet). The differences in the magnitude of odds ratio across states would suggest that state-level factors had a major influence on the attainment of digital skills.

The one way to specifically capture state-level policy on digital education would be to use the data from the Unified District Information on School Education (U-DISE) (GoI, 2019a), which gives us the number of schools with functional computers. Unfortunately, that data are not captured within the NSO. This variable is important because it conceptually captured another access point for digital devices, that is, students could access computers in schools. Learning at school could presumably help them get certifiable digital skills and therefore improve their employability.

6.2 Interaction effect of school digital infrastructure

Including state-wise number of functional computers in schools along with state-level dummy variables may result in multi-collinearity and we would have to drop one or the another in the regressions. Therefore, we constructed a median variable to overcome the issue. We computed the percentage of urban secondary and higher secondary schools with functional computers from U-DISE data (GoI, 2019a), and derived the median which was 47%. States with above-the-median percentages were given a value of one and those with below-the-median percentages got a value of zero: thus 12 states had a value of one and 12 had value of zero. This variable would henceforth be known as the threshold.

The proportion of schools with functional computers was more than the percentage of households with computers in urban areas for 16 states (Figure 4). The exceptions were Bihar, Uttar Pradesh, Odisha, Rajasthan, Assam, Telangana, Gujarat and the Northeast. Goa stood out with relatively higher access to digital devices both at home and in schools. Essentially, absence of access to digital devices at home may be mitigated by access at school, which may enable acquisition of digital skills. To capture this impact, we interacted the two dummy variables—the threshold variable (as defined above) with households having computers at home—and re-estimated the models. Essentially we were testing whether the odds ratio of students who belonged to row 1 in Table 7 was positive and statistically significant.

Figure 4

Source: Authors’ computations from GoI (2019a; 2019d).

Table 7

Interaction Variable of % of Urban Secondary and Senior Secondary Schools with a Computer Above the Median and Households with No Computer

S. No.	Threshold	Household Do Not Have Own/Access to a Computer	Interaction Variable
1.	1	1	1
2.	0	1	0
3.	1	0	0
4.	0	0	0

Source: The authors.

Table 8

(a) Interaction Equation: Odds Ratio on the Interaction Variable

		I		II		III
		Ability to Operate Computer		Ability to Browse Internet		Ability to Operate Computer and Browse Internet
		B	Odds Ratio	B	Odds Ratio	B	Odds Ratio
Exp (B) of households with computers		–1.91*** (0.003)	0.15**	–0.88*** [0.003]	0.42**	–1.65*** [0.003]	0.19**
Exp (B) of threshold		1.84*** (0.007)	6.27**	2.02*** [0.007]	7.57**	1.66*** [0.007]	5.27**
Exp (B) of interaction term		–0.02*** (0.005)	0.98**	–0.25*** [0.004]	0.78**	0.15*** [0.004]	1.17**
Overall percentage of cases correctly predicted by the model		78.9		80.2		77.8
Pseudo-R² (Nagelkerke)		0.441		0.448		0.445
(b) Odds Ratio of Focus Variables.
Focus Variable	Variables	Ability to Operate Computer		Ability to Browse Internet		Ability to Operate Computer and Browse Internet
Threshold = 1	Households with computers	6.27		7.57		5.27
Threshold = 1	Households without computers	6.14		5.88		6.15
Households without computers	Threshold = 1	0.15		0.32		0.22
Households without computers	Threshold = 0	0.15		0.42		0.19

Source: Authors’ computations from GoI (2019a; 2019d).

Note: (a) The model is based on urban students aged 14–21 years currently enrolled in Grades 9–12; (b) the model shows weighted results; (c) ***, ** and * indicate 1%, 5% and 10% levels of statistical significance, respectively and (d) the significance of the odds ratio is determined using the 95% confidence interval. If the interval includes one, then it is considered statistically insignificant.

Table 8 reported the key results. There was marginal improvement in the number of corrected predicted cases and pseudo-R² for the second and third regressions, but those stay the same for the first regression.

Both the threshold and interaction terms are reported in Table 8b. The odds of a student from a State with schools above the threshold value having both the ability to operate a computer and browse the Internet was 5.3 times higher than a student from a state which was below the threshold, ceteris paribus. The interaction term was significant in all the three regressions (Table 8a).

The sign on the interaction term, however, was negative in the first two regressions and positive in the third. The odds ratio (of no ownership/access to digital devices and living in an above median state) was higher for only one measure—the ability to operate a computer and browse the Internet (Table 8b). For the other two dependent variables odds ratios were lower, that is, the interaction has a negative impact. Therefore, there was some evidence that access to computers in schools may help overcome barriers of access to digital devices at home. This was especially important for the combination of skills—ability to operate a computer and browse the Internet. Presumably, this combination of skills would be what the employer would want! Last but not the least, public policy can influence the attainment of digital skills of secondary school students.

7. Summary and Policy Implications

The objective in this article was to document the employability of currently enrolled secondary and senior secondary students aged 14–21, specifically their functional digital skills. While there have been substantial improvements in functional digital skills between 2014 and 2017–2018, differences continue: across spatial (rural–urban, state), gender and socio-economic groups; and based on household ownership of and access to digital devices.

The econometric analysis suggested that access to devices was a critical constraint for students acquiring functional digital skills. Having more schools with functional computers could help overcome barriers of access to digital devices. This implied that State governments should improve digital infrastructure in all schools. Good data on this is fundamental to the success, so in addition, both the NSS and U-DISE need to strengthen their frameworks for capturing the scope and quality of digital skills among students. The limitations of data collected so far are that they do not allow us to assess the quality of skills attained, that is, whether young people actually have the skills schools are professing to impart and whether these skills will make them employable.

Footnotes

The authors declared no potential conflicts of interest with respect to the research, authorship and/or publication of this article.

Funding

The authors received no financial support for the research, authorship and/or publication of this article.

Notes

References

Annual Status of Education Report (ASER) (2018). ASER 2017 rural: Beyond basics. http://www.asercentre.org

Atasoy

(2011). ICT skills and employment opportunities [working paper #11-24]. http://www.netinst.org/Atasoy_11_24.pdf

Azam

, Chin

, & Prakash

(2010). The returns to English-language skills in India. Economic Development and Cultural Change, 61(2), 335–367.

Becker

G. S.

, & Lewis

H. G.

(1973). On the interaction between the quantity and quality of children. Journal of Political Economy, 81(2), S279–S288.

Burns

, & Gottschalk

(Eds.). (2019). Educating 21st century children: Emotional well-being in the digital age, educational research and innovation. OECD Publishing. https://doi.org/10.1787/b7f33425-en

Falck

, Heimisch

, & Wiederhold

(2016). Returns to ICT skills [OECD education working papers, no. 134]. OECD Publishing. https://dx-doi-org.web.bisu.edu.cn/10.1787/5jlzfl2p5rzq-en

Garrido

, Sullivan

, & Gordon

(2012). Understanding the links between ICT skills training and employability: An analytical framework. Information Technologies and International Development, 8(2), 17–32. https://itidjournal.org/index.php/itid/article/view/895

Government of India (GoI) (2010). Education in India: 2007–08 participation and expenditure. NSS 64th Round (July 2007–June 2008) [report no. 532(64/25.2/1)]. Ministry of Statistics and Programme Implementation, New Delhi.

Government of India (GoI) (2016). Education in India: 71^st Round (January-June 2014). Ministry of Statistics and Programme Implementation, New Delhi. www.mospi.gov.in

10.

Government of India (GoI) (2019a). Dashboard. Department of School Education & Literacy, Ministry of Human Resource Development. UDISE+. http://dashboard.udiseplus.gov.in/#!/reports

11.

Government of India (GoI) (2019b). All-India survey on higher education 2018–19. Department of Higher Education, Ministry of Human Resources and Development. www.aishe.nic.in

12.

Government of India (GoI) (2019c). Periodic labour force survey (July 2017–June 2018). Ministry of Statistics and Programme Implementation, New Delhi. www.mospi.gov.in

13.

Government of India (GoI) (2019d). Key indicators of household social consumption on education in India: 75th Round (July 2017-June 2018). Ministry of Statistics and Programme Implementation, New Delhi. www.mospi.gov.in

14.

Government of India (GoI) (2020). National Education Policy 2020. Ministry of Human Resource and Development, New Delhi. https://www.mhrd.gov.in/sites/upload_files/mhrd/files/NEP_Final_English_0.pdf

15.

Greene

W. H.

(2000). Econometric analysis (4th ed.). Prentice Hall.

16.

IBM Knowledge Centre website Pseudo R-Squared Measures. https://www.ibm.com/support/knowledgecenter/pl/SSLVMB_24.0.0/spss/tutorials/plum_germcr_rsquare.html

17.

Kugler

, & Kumar

(2017). Preference for boys, family size, and educational attainment in India. Demography, 54, 835–859.

18.

Mudassir

I. U.

, & Abubakar

B. A.

(2015). The impact of parents’ occupation on academic performance of secondary school students in Kuala Terengganu. Multilingual Academic Journal of Education and Social Sciences, 3(1), 112–120.

19.

Mukherjee

, Mazar

, Aggarwal

, & Kumar

(2016). Exclusion from digital infrastructure and access. Digital Empowerment Foundation, New Delhi. www.defindia.org/publication-2

20.

Myroniuk

T. W.

, Vanneman

, & Desai

(2017). Getting a child through secondary school and to college in India: The role of household social capital. Sociology of Development (Oakl), 3(1), 24–46.

21.

National Council of Applied Economic Research (NCAER) (2018). Skilling India: No time to lose. http://www.ncaer.org/publication_details.php?pID=300

22.

National Council of Educational Research and Training (NCERT) website. http://www.ncert.nic.in/index.html

23.

National Council of Educational Research and Training (NCERT) (2015). A summary of National Achievement Survey: Class X. National Council of Educational Research and Training, New Delhi.

24.

Ocansey

T. S.

, & Sharma

(2020). Maharashtra digital schools survey findings report [ICT India working paper # 24]. Center for Sustainable Development, Earth Institute, Columbia University, New York. https://csd.columbia.edu/sites/default/files/content/docs/ICT%20India/Papers/ICT_India_Working_Paper_24.pdf

25.

Office of the Registrar General & Census Commissioner India (2011). HH-12 number of households availing banking services and number of households having each of the specified assets. https://censusindia.gov.in/2011census/Hlo-series/HH12.html

26.

Rani

P. G.

, Shree

, & Shukla

(2019). Return to skills in India: The role of digital access and usage. Indian Journal of Human Development, 13(3), 254–277.

27.

Sunder

(2019, May 20). It runs in the family: Parents’ education and child learning outcomes. Ideas for India. https://www.ideasforindia.in/topics/human-development/it-runs-in-the-family-parents-education-and-child-learning-outcomes.html

28.

Telecom Regulatory Authority of India (TRAI) (2019, September 25). Indian Telecom Sector 2018 (3rd ed.). https://trai.gov.in/release-publication/reports/performance-indicators-reports. New Delhi, India.

29.

Thakur

G. K.

(2014). ICT and digital divide in Indian school system. International Journal of Interdisciplinary and Multidisciplinary Studies (IJIMS), 2(2), 34–38.

30.

United Nations Educational, Scientific and Cultural, Organization (UNESCO) and Intel (2017, September). Working Group on Education: Digital skills for life and work. https://www.skillsforemployment.org/KSP/en/Details/?dn=WCMSTEST4_193296

31.

United Nations Educational, Scientific and Cultural, Organization (UNESCO) (2018, March 15). Digital skills critical for jobs and social inclusion. https://en.unesco.org/news/digital-skills-critical-jobs-and-social-inclusion

32.

World Development Indicators (WDI) website. (n.d.). World Development Indicators. World Bank, Washington D.C. http://data.worldbank.org/data-catalog/world-development-indicators

33.

Young Lives (2017). Education and Learning: Preliminary Findings from the 2016 Young Lives Survey (Round 5): United Andhra Pradesh. https://www.younglives-india.org/sites/www.younglives-india.org/files/2017-09/INDIA-Education%26Learning-Factsheet-Aug17-Proof03.pdf