The Degree Is Not Enough: Gendered Perceptions of Employability and Self-Directed Skilling Among Generation Z Graduates in Kerala

Higher education systems everywhere are being asked to answer a question they were not originally built to answer: not simply whether a graduate is educated, but whether a graduate is deployable. The knowledge economy has reorganised employer expectations around adaptable, practical competence rather than the accumulation of credentials, and the pace of technological change — automation, data-driven work, and now generative artificial intelligence — has made the useful shelf-life of a fixed curriculum shorter than the time it takes to complete a degree. Employers describe graduates who can recite theory but hesitate at a spreadsheet, a client call, or an unfamiliar tool, and the resulting vocabulary of 'skill gap' and 'employability' has migrated from human-resource departments into the language of higher-education policy itself.

India's version of this conversation carries its own weight of figures. Multiple industry assessments converge on the finding that fewer than half of graduating students are considered employable in roles matched to their qualifications, and the reasons cited are strikingly consistent: weak communication, limited problem-solving ability, and a shortage of practical, technology-linked competence rather than any deficit in formal subject knowledge (Mercer | Mettl, 2025). The gap is not simply about the absence of jobs; it is about a mismatch between what degree programmes certify and what a modern workplace actually requires, a mismatch that has proven resistant to two decades of expansion in gross enrolment.

Kerala occupies a distinctive and somewhat paradoxical position within this national picture. The state's educational achievements — near-universal literacy, high secondary and tertiary enrolment, and a strong cultural premium on formal education — sit alongside a labour market that has struggled to absorb its own graduates, pushing a significant share of educated youth toward migration, informal work, or prolonged underemployment. George and Menon (2023) describe this as a withering of confidence in the state's higher-education sector, driven by skill mismatches, stagnant curricula, and a widening gap between institutional output and industry need. Where earlier generations of Kerala graduates faced a shortage of jobs, the present generation increasingly reports a shortage of fit: credentials that satisfy the syllabus but not the employer.

What remains comparatively under-documented is how this perceived gap is experienced differently by male and female graduates. International research on self-perceived employability has generally found that women report lower confidence in their employability than men, even when their measured skills or academic performance are equivalent (Bennett et al., 2021; Wen et al., 2023). This asymmetry has typically been attributed to social and institutional conditioning that discourages women from claiming competence as readily as their male peers, rather than to any underlying difference in ability. Whether this pattern holds, is absent, or is reversed among Generation Z graduates in a specific regional context such as Kerala has not been directly established, and this is the empirical gap the present study addresses.

A second, more recent dimension of the skill-acquisition question concerns the role of generative artificial intelligence. Existing work on Generation Z's relationship with tools such as ChatGPT describes a cohort markedly more comfortable adopting these tools for academic and skill-building purposes than the generations that teach them, often treating generative AI as a routine study companion rather than a novelty (Lee et al., 2023). Where a formal degree programme is perceived as slow to update, generative AI and other freely available digital resources may increasingly substitute for classroom instruction as the practical channel through which graduates build the skills they believe employers want. This study treats that possibility as an empirical question rather than an assumption.

Accordingly, this study investigates four linked questions using survey data collected from a Generation Z sample in Kerala, disaggregated by gender throughout: (1) how sufficient respondents perceive their formal college degree to be for securing an immediate, industry-ready job; (2) what respondents identify as the single greatest flaw in the state's higher-education system; (3) through which channels respondents primarily acquire modern, industry-relevant skills, including the role of generative AI; and (4) which credential or resource respondents value most highly when trying to make their résumé employable.

This study relies exclusively on a structured, closed-ended survey dataset supplied for analysis; no additional data collection, secondary dataset, or supplementary source of primary evidence was used. The instrument presented each item bilingually, in English and Malayalam, and required respondents to select a single response from a fixed set of categories for each question. Five substantive items form the basis of this analysis: perceived sufficiency of the formal degree for immediate, industry-ready employment; the single greatest perceived flaw in Kerala's higher-education system; the primary channel through which respondents acquire modern, industry-relevant skills; the frequency of use of generative AI tools for academic work; and the educational resource most valued for making a résumé or CV employable.

Responses were disaggregated by self-reported gender (female and male) and are additionally reported as pooled totals across the full sample. Because only response-category percentages were supplied, the analytical strategy adopted is descriptive: percentage distributions are compared across gender and against the pooled sample to identify patterns, contrasts, and areas of convergence or divergence. No inferential statistical procedure — significance testing, regression, or factor analysis — was available in the supplied data, and none is claimed or implied in the analysis that follows.

Data limitation disclosed by the author: one item, concerning the channel through which respondents acquire industry-relevant skills, exhibited an apparent inconsistency in the male-disaggregated figures — the categories and percentages recorded for male respondents numerically mirror those recorded for a separate item on generative AI usage frequency. This is treated as a probable labelling or chart-export artefact in the source material rather than as substantive data, and the male-specific distribution for that item is consequently not reported. The female-disaggregated and pooled totals for the same item, which do not show this anomaly, are reported and discussed normally.

All findings are reported as observed percentage distributions rounded to one decimal place, as supplied in the source material. Interpretation of these patterns, and their relationship to existing scholarship, is reserved for the Discussion section; the Results section that follows is restricted to description.

Perceived sufficiency of the degree. When asked whether their formal college degree could secure them an immediate, industry-ready job, female and male respondents diverged sharply, and in opposite directions to what a simple deficit narrative might predict. Female respondents were considerably more likely than male respondents to describe their degree as fully sufficient (40.4% versus 22.1%), while male respondents were roughly twice as likely as female respondents to describe their degree as completely insufficient (32.4% versus 15.6%). Both groups converged on the middle category, describing their degree as only partially sufficient (44.0% of female respondents, 45.5% of male respondents). Pooled across the sample, the modal response indicated partial sufficiency (45.1%).

Perceived flaw in the higher-education system. An outdated theoretical syllabus with no practical training was the most frequently selected flaw across genders, but the gap between genders was substantial: 62.4% of male respondents selected this option compared with 40.4% of female respondents. Female respondents distributed their remaining responses more evenly, with 32.1% citing a complete lack of corporate placements and 19.3% citing systemic delays in examinations and results, compared with 21.4% and 7.9% respectively among male respondents. A lack of qualified, industry-experienced professors was the least selected flaw for both genders (8.3% each). In the pooled sample, the outdated syllabus was the dominant response (56.4%), followed by the absence of placements (24.3%).

Channels of skill acquisition. Female respondents reported acquiring modern, industry-relevant skills predominantly through YouTube and free open-source material (68.8%), followed by reliance strictly on formal college classes (22.0%) and paid online certification courses (7.3%); government training schemes such as ASAP Kerala or K-DISC accounted for a negligible share (1.8%). In the pooled sample, YouTube and free online material remained the leading channel (50.1%), with formal college classes cited by roughly a third of respondents (32.8%) and paid certification courses by 15.3%. As noted in the Methodology, the male-disaggregated figures for this specific item could not be reliably distinguished from those of a separate item in the source material and are therefore not reported.

Frequency of generative AI use. A majority of both female (48.6%) and male (61.4%) respondents reported using generative AI tools daily or for almost all assignments, and a further 44.0% of female and 35.5% of male respondents reported occasional use for reference. Reporting no use at all was rare in both groups (0.9% of female respondents; negligible among male respondents). Pooled across the sample, 57.9% of respondents reported daily or near-daily use, indicating that generative AI has become a routine rather than an exceptional part of academic practice for this cohort.

Most valued employability resource. Asked which resource they valued most for making their résumé or CV employable, the largest share of both female (35.8%) and male (41.2%) respondents selected the prestige and percentage of their university degree, ahead of a strong portfolio of independent practical projects (24.8% female, 30.4% male) and global skill certifications from providers such as Google or AWS (18.3% female, 20.4% male). Letters of recommendation from internships were the least valued resource for male respondents (8.0%) and the second-least valued for female respondents (21.1%). In the pooled sample, the university degree remained the most valued single resource (39.6%), followed by a portfolio of projects (29.1%).

The most striking pattern in these findings runs against the grain of much existing employability research. International literature has generally found that women report lower self-perceived employability than men despite comparable or superior academic performance, attributing this asymmetry to social conditioning that discourages women from claiming competence (Bennett et al., 2021; Wen et al., 2023). The present data show the opposite direction for the specific question of degree sufficiency: female respondents were markedly more confident that their degree was fully sufficient for immediate, industry-ready work, while male respondents were both more likely to view their degree as inadequate and more likely to blame the syllabus's outdated theoretical content specifically. One plausible reading, consistent with rather than proven by the data, is that male respondents in this sample may hold more exacting expectations of what a job-ready credential should deliver, particularly in technical or competitive fields where the gap between coursework and employer expectation is most visible. This divergence should be treated as an observation requiring further, targeted investigation rather than a settled explanation, since the dataset does not disaggregate respondents by academic discipline.

The skill-acquisition findings sit comfortably alongside the broader Indian evidence base, which consistently identifies practical and digital competence, rather than subject knowledge, as the primary constraint on employability (Mercer | Mettl, 2025). What the present data add is a more granular picture of where that practical competence is actually being built: not principally in the classroom, but through free digital material and, increasingly, generative AI. That over half of the pooled sample reports daily or near-daily generative AI use for academic work is consistent with recent findings that Generation Z students adopt such tools more readily, and with less hesitation, than the generations who teach them (Lee et al., 2023). Read alongside the finding that formal college classes are a minority channel for skill acquisition, this suggests a quiet redistribution of where meaningful learning is perceived to occur — a redistribution the curriculum has evidently not yet caught up with, consistent with George and Menon's (2023) account of a Kerala higher-education sector whose pace of reform lags behind labour-market change.

The résumé-value finding complicates a simple 'credential is dead' narrative, however. Despite widespread scepticism about degree sufficiency, respondents of both genders still ranked the prestige and percentage of their university degree above independent projects, certifications, or recommendation letters as the resource that makes them employable. This is not necessarily inconsistent with the earlier findings: respondents can simultaneously doubt that a degree delivers practical readiness and still believe that the degree functions as the signal employers use to sort candidates. The two beliefs describe different functions of a credential — one concerning capability, the other concerning legitimacy — and graduates may reasonably hold both at once. A similar distinction appears in comparative work on engineering and management curricula in India (Murari, 2026).

Several limitations of this analysis follow directly from the constraints of the supplied dataset. The absence of sample size, institutional affiliation, and discipline-level detail means the patterns reported here cannot be generalised beyond the sample surveyed, and the apparent labelling anomaly in the male skill-acquisition figures further restricts what can be said about that single item. The descriptive design also means that none of the gender differences reported can be characterised as statistically significant, since no inferential test was available; the differences are notable in magnitude but should be read as suggestive rather than confirmatory.

This study examined how a Generation Z sample in Kerala perceives the sufficiency of their formal education for industry-ready employment, disaggregated by gender, using a descriptive analysis of survey data. Female respondents expressed comparatively greater confidence in their degree's sufficiency, while male respondents were more critical and specifically attributed the state's higher-education shortfalls to an outdated, impractical syllabus. Across genders, self-directed digital learning and generative AI use have become dominant channels through which respondents report building industry-relevant skills, even as the university degree retains its perceived value as a credentialing signal for employability.

These findings offer preliminary, descriptive evidence rather than a generalisable causal account, and they should be read as an invitation to further, more granular research — particularly research that disaggregates by academic discipline and institution. For policymakers, curriculum designers, and institutions engaged in the kind of evidence review this journal is intended to support, the clearest signal from this dataset is that the gap graduates describe is not primarily a gap in subject knowledge, but in the practical, self-directed, and increasingly AI-mediated skilling that formal coursework has not yet absorbed.